Replacing nodes
Instructions for how to safely replace a node in an OpenShift Data Foundation cluster.
Abstract
Making open source more inclusive
Red Hat is committed to replacing problematic language in our code, documentation, and web properties. We are beginning with these four terms: master, slave, blacklist, and whitelist. Because of the enormity of this endeavor, these changes will be implemented gradually over several upcoming releases. For more details, see our CTO Chris Wright’s message.
Providing feedback on Red Hat documentation
We appreciate your input on our documentation. Do let us know how we can make it better.
To give feedback, create a Bugzilla ticket:
- Go to the Bugzilla website.
- In the Component section, choose documentation.
- Fill in the Description field with your suggestion for improvement. Include a link to the relevant part(s) of documentation.
- Click Submit Bug.
Preface
For OpenShift Data Foundation, node replacement can be performed proactively for an operational node and reactively for a failed node for the following deployments:
For Amazon Web Services (AWS)
- User-provisioned infrastructure
- Installer-provisioned infrastructure
For VMware
- User-provisioned infrastructure
- Installer-provisioned infrastructure
For Microsoft Azure
- Installer-provisioned infrastructure
For local storage devices
- Bare metal
- VMware
- IBM Power
- For replacing your storage nodes in external mode, see Red Hat Ceph Storage documentation.
Chapter 1. OpenShift Data Foundation deployed using dynamic devices
1.1. OpenShift Data Foundation deployed on AWS
To replace an operational node, see:
To replace a failed node, see:
1.1.1. Replacing an operational AWS node on user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure and resources to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
When replacing an AWS node on user-provisioned infrastructure, the new node needs to be created in the same AWS zone as the original node.
Procedure
- Identify the node that you need to replace.
Mark the node as unschedulable:
$ oc adm cordon <node_name>
<node_name>
- Specify the name of node that you need to replace.
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
Delete the node:
$ oc delete nodes <node_name>
- Create a new Amazon Web Service (AWS) machine instance with the required infrastructure. See Platform requirements.
- Create a new OpenShift Container Platform node using the new AWS machine instance.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.1.2. Replacing an operational AWS node on installer-provisioned infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the node that you need to replace. Take a note of its Machine Name.
Mark the node as unschedulable:
$ oc adm cordon <node_name>
<node_name>
- Specify the name of node that you need to replace.
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Machines. Search for the required machine.
- Besides the required machine, click Action menu (⋮) → Delete Machine.
- Click Delete to confirm that the machine is deleted. A new machine is automatically created.
Wait for the new machine to start and transition into Running state.
ImportantThis activity might take at least 5 - 10 minutes or more.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.1.3. Replacing a failed AWS node on user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure and resources to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
- Identify the Amazon Web Service (AWS) machine instance of the node that you need to replace.
- Log in to AWS, and terminate the AWS machine instance that you identified.
- Create a new AWS machine instance with the required infrastructure. See Platform requirements.
- Create a new OpenShift Container Platform node using the new AWS machine instance.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Execute the following command to apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.1.4. Replacing a failed AWS node on installer-provisioned infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the faulty node, and click on its Machine Name.
- Click Actions → Edit Annotations, and click Add More.
-
Add
machine.openshift.io/exclude-node-draining
, and click Save. - Click Actions → Delete Machine, and click Delete.
A new machine is automatically created, wait for new machine to start.
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
- Optional: If the failed Amazon Web Service (AWS) instance is not removed automatically, terminate the instance from the AWS console.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.2. OpenShift Data Foundation deployed on VMware
To replace an operational node, see:
To replace a failed node, see:
1.2.1. Replacing an operational VMware node on user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure and resources to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
- Identify the node and its Virtual Machine (VM) that you need replace.
Mark the node as unschedulable:
$ oc adm cordon <node_name>
<node_name>
- Specify the name of node that you need to replace.
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
Delete the node:
$ oc delete nodes <node_name>
Log in to VMware vSphere, and terminate the VM that you identified:
ImportantDelete the VM only from the inventory and not from the disk.
- Create a new VM on VMware vSphere with the required infrastructure. See Platform requirements.
- Create a new OpenShift Container Platform worker node using the new VM.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.2.2. Replacing an operational VMware node on installer-provisioned infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the node that you need to replace. Take a note of its Machine Name.
Mark the node as unschedulable:
$ oc adm cordon <node_name>
<node_name>
- Specify the name of node that you need to replace.
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Machines. Search for the required machine.
- Besides the required machine, click Action menu (⋮) → Delete Machine.
- Click Delete to confirm the machine is deleted. A new machine is automatically created.
Wait for the new machine to start and transition into Running state.
ImportantThis activity might take at least 5 - 10 minutes or more.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.2.3. Replacing a failed VMware node on user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure and resources to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
- Identify the node and its Virtual Machine (VM) that you need to replace.
Delete the node:
$ oc delete nodes <node_name>
<node_name>
- Specify the name of node that you need to replace.
Log in to VMware vSphere and terminate the VM that you identified.
ImportantDelete the VM only from the inventory and not from the disk.
- Create a new VM on VMware vSphere with the required infrastructure. See Platform requirements.
- Create a new OpenShift Container Platform worker node using the new VM.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.2.4. Replacing a failed VMware node on installer-provisioned infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the faulty node, and click on its Machine Name.
- Click Actions → Edit Annotations, and click Add More.
-
Add
machine.openshift.io/exclude-node-draining
, and click Save. - Click Actions → Delete Machine, and click Delete.
A new machine is automatically created. Wait for te new machine to start.
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
- Optional: If the failed Virtual Machine (VM) is not removed automatically, terminate the VM from VMware vSphere.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.3. OpenShift Data Foundation deployed on Microsoft Azure
1.3.1. Replacing operational nodes on Azure installer-provisioned infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the node that you need to replace. Take a note of its Machine Name.
Mark the node as unschedulable:
$ oc adm cordon <node_name>
<node_name>
- Specify the name of node that you need to replace.
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Machines. Search for the required machine.
- Besides the required machine, click the Action menu (⋮) → Delete Machine.
- Click Delete to confirm the machine is deleted. A new machine is automatically created.
Wait for the new machine to start and transition into Running state.
ImportantThis activity might take at least 5 - 10 minutes or more.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Execute the following command to apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads→ Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.3.2. Replacing failed nodes on Azure installer-provisioned infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the faulty node, and click on its Machine Name.
- Click Actions → Edit Annotations, and click Add More.
-
Add
machine.openshift.io/exclude-node-draining
, and click Save. - Click Actions → Delete Machine, and click Delete.
A new machine is automatically created. Wait for the new machine to start.
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
- Optional: If the failed Azure instance is not removed automatically, terminate the instance from the Azure console.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that new the Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.4. OpenShift Data Foundation deployed on Google cloud
1.4.1. Replacing operational nodes on Google Cloud installer-provisioned infrastructure
Procedure
- Log in to OpenShift Web Console and click Compute → Nodes.
- Identify the node that needs to be replaced. Take a note of its Machine Name.
Mark the node as unschedulable using the following command:
$ oc adm cordon <node_name>
Drain the node using the following command:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
ImportantThis activity may take at least 5-10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when the new node is labeled and functional.
- Click Compute → Machines. Search for the required machine.
- Besides the required machine, click the Action menu (⋮) → Delete Machine.
- Click Delete to confirm the machine deletion. A new machine is automatically created.
Wait for new machine to start and transition into Running state.
ImportantThis activity may take at least 5-10 minutes or more.
- Click Compute → Nodes, confirm if the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From User interface
- For the new node, click Action Menu (⋮) → Edit Labels
-
Add
cluster.ocs.openshift.io/openshift-storage
and click Save.
- From Command line interface
Execute the following command to apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that the new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
1.4.2. Replacing failed nodes on Google Cloud installer-provisioned infrastructure
Procedure
- Log in to OpenShift Web Console and click Compute → Nodes.
- Identify the faulty node and click on its Machine Name.
- Click Actions → Edit Annotations, and click Add More.
-
Add
machine.openshift.io/exclude-node-draining
and click Save. - Click Actions → Delete Machine, and click Delete.
A new machine is automatically created, wait for new machine to start.
ImportantThis activity may take at least 5-10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when the new node is labeled and functional.
- Click Compute → Nodes, confirm if the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the web user interface
- For the new node, click Action Menu (⋮) → Edit Labels
-
Add
cluster.ocs.openshift.io/openshift-storage
and click Save.
- From the command line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
- Optional: If the failed Google Cloud instance is not removed automatically, terminate the instance from Google Cloud console.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that new the Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
Chapter 2. OpenShift Data Foundation deployed using local storage devices
2.1. Replacing storage nodes on bare metal infrastructure
- To replace an operational node, see Section 2.1.1, “Replacing an operational node on bare metal user-provisioned infrastructure”.
- To replace a failed node, see Section 2.1.2, “Replacing a failed node on bare metal user-provisioned infrastructure”.
2.1.1. Replacing an operational node on bare metal user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure, resources, and disks to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
Identify the node, and get the labels on the node that you need to replace:
$ oc get nodes --show-labels | grep <node_name>
<node_name>
- Specify the name of node that you need to replace.
Identify the monitor pod (if any), and OSDs that are running in the node that you need to replace:
$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon <node_name>
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
Delete the node:
$ oc delete node <node_name>
Get a new bare-metal machine with the required infrastructure. See Installing on bare metal.
ImportantFor information about how to replace a master node when you have installed OpenShift Data Foundation on a three-node OpenShift compact bare-metal cluster, see the Backup and Restore guide in the OpenShift Container Platform documentation.
- Create a new OpenShift Container Platform node using the new bare-metal machine.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Identify the namespace where OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a new worker node to the
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumediscovery auto-discover-devices
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed, andnewnode.example.com
is the new node.Determine the
localVolumeSet
to edit:# oc get -n $local_storage_project localvolumeset
Example output:
NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumeset localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In the this example,
server3.example.com
is removed andnewnode.example.com
is the new node.
Verify that the new
localblock
Persistent Volume (PV) is available:$oc get pv | grep localblock | grep Available
Example output:
local-pv-551d950 512Gi RWO Delete Available localblock 26s
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix.You can add comma separated OSD IDs in the command to remove more than one OSD, for example,
FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails, and the pod is not in the expectedCompleted
state, check the pod logs for further debugging:For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Identify the Persistent Volume (PV) associated with the Persistent Volume Claim (PVC):
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-d6bf175b 1490Gi RWO Delete Released openshift-storage/ocs-deviceset-0-data-0-6c5pw localblock 2d22h compute-1
If there is a PV in
Released
state, delete it:# oc delete pv <persistent_volume>
For example:
# oc delete pv local-pv-d6bf175b
Example output:
persistentvolume "local-pv-d9c5cbd6" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in
Running
state:-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in
Running
state.Ensure that the new incremental
mon
is created, and is in theRunning
state:$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-a-cd575c89b-b6k66 2/2 Running 0 38m rook-ceph-mon-b-6776bc469b-tzzt8 2/2 Running 0 38m rook-ceph-mon-d-5ff5d488b5-7v8xh 2/2 Running 0 4m8s
OSD and monitor pod might take several minutes to get to the
Running
state.Verify that new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.1.2. Replacing a failed node on bare metal user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure, resources, and disks to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
Identify the node, and get the labels on the node that you need to replace:
$ oc get nodes --show-labels | grep <node_name>
<node_name>
- Specify the name of node that you need to replace.
Identify the monitor pod (if any), and OSDs that are running in the node that you need to replace:
$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon <node_name>
Remove the pods which are in
Terminating
state:$ oc get pods -A -o wide | grep -i <node_name> | awk '{if ($4 == "Terminating") system ("oc -n " $1 " delete pods " $2 " --grace-period=0 " " --force ")}'
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
Delete the node:
$ oc delete node <node_name>
Get a new bare-metal machine with the required infrastructure. See Installing on bare metal.
ImportantFor information about how to replace a master node when you have installed OpenShift Data Foundation on a three-node OpenShift compact bare-metal cluster, see the Backup and Restore guide in the OpenShift Container Platform documentation.
- Create a new OpenShift Container Platform node using the new bare-metal machine.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Identify the namespace where OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a new worker node to the
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumediscovery auto-discover-devices
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed, andnewnode.example.com
is the new node.Determine the
localVolumeSet
to edit:# oc get -n $local_storage_project localvolumeset
Example output:
NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumeset localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In the this example,
server3.example.com
is removed andnewnode.example.com
is the new node.
Verify that the new
localblock
Persistent Volume (PV) is available:$oc get pv | grep localblock | grep Available
Example output:
local-pv-551d950 512Gi RWO Delete Available localblock 26s
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix.You can add comma separated OSD IDs in the command to remove more than one OSD, for example,
FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails, and the pod is not in the expectedCompleted
state, check the pod logs for further debugging:For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Identify the Persistent Volume (PV) associated with the Persistent Volume Claim (PVC):
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-d6bf175b 1490Gi RWO Delete Released openshift-storage/ocs-deviceset-0-data-0-6c5pw localblock 2d22h compute-1
If there is a PV in
Released
state, delete it:# oc delete pv <persistent_volume>
For example:
# oc delete pv local-pv-d6bf175b
Example output:
persistentvolume "local-pv-d9c5cbd6" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in
Running
state:-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in
Running
state.Ensure that the new incremental
mon
is created, and is in theRunning
state:$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-a-cd575c89b-b6k66 2/2 Running 0 38m rook-ceph-mon-b-6776bc469b-tzzt8 2/2 Running 0 38m rook-ceph-mon-d-5ff5d488b5-7v8xh 2/2 Running 0 4m8s
OSD and monitor pod might take several minutes to get to the
Running
state.Verify that new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.2. Replacing storage nodes on IBM Z or IBM® LinuxONE infrastructure
You can choose one of the following procedures to replace storage nodes:
2.2.1. Replacing operational nodes on IBM Z or IBM® LinuxONE infrastructure
Use this procedure to replace an operational node on IBM Z or IBM® LinuxONE infrastructure.
Procedure
Identify the node and get labels on the node to be replaced. Make a note of the rack label.
$ oc get nodes --show-labels | grep <node_name>
Identify the mon (if any) and object storage device (OSD) pods that are running in the node to be replaced.
$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods identified in the previous step.
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage $ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage $ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the nodes as unschedulable.
$ oc adm cordon <node_name>
Remove the pods which are in the
Terminating
state.$ oc get pods -A -o wide | grep -i <node_name> | awk '{if ($4 == "Terminating") system ("oc -n " $1 " delete pods " $2 " --grace-period=0 " " --force ")}'
Drain the node.
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
Delete the node.
$ oc delete node <node_name>
- Get a new IBM Z storage node as a replacement.
Check for certificate signing requests (CSRs) related to OpenShift Data Foundation that are in
Pending
state:$ oc get csr
Approve all required OpenShift Data Foundation CSRs for the new node:
$ oc adm certificate approve <Certificate_Name>
- Click Compute → Nodes in OpenShift Web Console, confirm if the new node is in Ready state.
Apply the
openshift-storage
label to the new node using any one of the following:- From User interface
- For the new node, click Action Menu (⋮) → Edit Labels
-
Add
cluster.ocs.openshift.io/openshift-storage
and click Save.
- From Command line interface
- Execute the following command to apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
Add a new worker node to
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node and remove the failed node.# oc edit -n local-storage-project localvolumediscovery auto-discover-devices [...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In the above example,
server3.example.com
was removed andnewnode.example.com
is the new node.Determine which
localVolumeSet
to edit.Replace local-storage-project in the following commands with the name of your local storage project. The default project name is
openshift-local-storage
in OpenShift Data Foundation 4.6 and later. Previous versions uselocal-storage
by default.# oc get -n local-storage-project localvolumeset NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node and remove the failed node.# oc edit -n local-storage-project localvolumeset localblock [...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In the above example,
server3.example.com
was removed andnewnode.example.com
is the new node.
Verify that the new
localblock
PV is available.$ oc get pv | grep localblock CAPA- ACCESS RECLAIM STORAGE NAME CITY MODES POLICY STATUS CLAIM CLASS AGE local-pv- 931Gi RWO Delete Bound openshift-storage/ localblock 25h 3e8964d3 ocs-deviceset-2-0 -79j94 local-pv- 931Gi RWO Delete Bound openshift-storage/ localblock 25h 414755e0 ocs-deviceset-1-0 -959rp local-pv- 931Gi RWO Delete Available localblock 3m24s b481410 local-pv- 931Gi RWO Delete Bound openshift-storage/ localblock 25h d9c5cbd6 ocs-deviceset-0-0 -nvs68
Change to the
openshift-storage
project.$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required.
Identify the PVC as afterwards we need to delete PV associated with that specific PVC.
$ osd_id_to_remove=1 $ oc get -n openshift-storage -o yaml deployment rook-ceph-osd-${osd_id_to_remove} | grep ceph.rook.io/pvc
where,
osd_id_to_remove
is the integer in the pod name immediately after therook-ceph-osd prefix
. In this example, the deployment name isrook-ceph-osd-1
.Example output:
ceph.rook.io/pvc: ocs-deviceset-localblock-0-data-0-g2mmc ceph.rook.io/pvc: ocs-deviceset-localblock-0-data-0-g2mmc
In this example, the PVC name is
ocs-deviceset-localblock-0-data-0-g2mmc
.Remove the failed OSD from the cluster.
$ oc process -n openshift-storage ocs-osd-removal -p FAILED_OSD_IDS=${osd_id_to_remove} |oc create -f -
You can remove more than one OSD by adding comma separated OSD IDs in the command. (For example: FAILED_OSD_IDS=0,1,2)
WarningThis step results in OSD being completely removed from the cluster. Ensure that the correct value of
osd_id_to_remove
is provided.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-osd_id_to_remove -n openshift-storage
NoteIf
ocs-osd-removal
fails and the pod is not in the expectedCompleted
state, check the pod logs for further debugging. For example:# oc logs -l job-name=ocs-osd-removal-osd_id_to_remove -n openshift-storage --tail=-1
It may be necessary to manually cleanup the removed OSD as follows:
ceph osd crush remove osd.osd_id_to_remove ceph osd rm osd_id_to_remove ceph auth del osd.osd_id_to_remove ceph osd crush rm osd_id_to_remove
Delete the PV associated with the failed node.
Identify the PV associated with the PVC.
The PVC name must be identical to the name that is obtained while removing the failed OSD from the cluster.
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released local-pv-5c9b8982 500Gi RWO Delete Released openshift-storage/ocs-deviceset-localblock-0-data-0-g2mmc localblock 24h worker-0
If there is a PV in
Released
state, delete it.# oc delete pv <persistent-volume>
For example:
# oc delete pv local-pv-5c9b8982 persistentvolume "local-pv-5c9b8982" deleted
Identify the
crashcollector
pod deployment.$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it.$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal
job.# oc delete job ocs-osd-removal-${osd_id_to_remove}
Example output:
job.batch "ocs-osd-removal-0" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If data encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.2.2. Replacing failed nodes on IBM Z or IBM® LinuxONE infrastructure
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the faulty node, and click on its Machine Name.
- Click Actions → Edit Annotations, and click Add More.
-
Add
machine.openshift.io/exclude-node-draining
, and click Save. - Click Actions → Delete Machine, and click Delete.
A new machine is automatically created. Wait for new machine to start.
ImportantThis activity might take at least 5 - 10 minutes or more. Ceph errors generated during this period are temporary and are automatically resolved when you label the new node, and it is functional.
- Click Compute → Nodes. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= | cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
- Verify that all the other required OpenShift Data Foundation pods are in Running state.
Verify that new Object Storage Device (OSD) pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If data encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.3. Replacing storage nodes on IBM Power infrastructure
For OpenShift Data Foundation, you can perform node replacement proactively for an operational node, and reactively for a failed node, for the deployments related to IBM Power.
2.3.1. Replacing an operational or failed storage node on IBM Power
Prerequisites
- Ensure that the replacement nodes are configured with the similar infrastructure and resources to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
Identify the node, and get the labels on the node that you need to replace:
$ oc get nodes --show-labels | grep <node_name>
<node_name>
- Specify the name of node that you need to replace.
Identify the
mon
(if any), and Object Storage Device (OSD) pods that are running in the node that you need to replace:$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-a --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-1 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon <node_name>
Remove the pods which are in
Terminating
state:$ oc get pods -A -o wide | grep -i <node_name> | awk '{if ($4 == "Terminating") system ("oc -n " $1 " delete pods " $2 " --grace-period=0 " " --force ")}'
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
Delete the node:
$ oc delete node <node_name>
- Get a new IBM Power machine with the required infrastructure. See Installing a cluster on IBM Power.
- Create a new OpenShift Container Platform node using the new IBM Power machine.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=''
<new_node_name>
- Specify the name of the new node.
Identify the namespace where OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a newly added worker node to the
localVolume
.Determine the
localVolume
you need to edit:# oc get -n $local_storage_project localvolume
Example output:
NAME AGE localblock 25h
Update the
localVolume
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolume localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: #- worker-0 - worker-1 - worker-2 - worker-3 [...]
Remember to save before exiting the editor.
In the this example,
worker-0
is removed andworker-3
is the new node.
Verify that the new
localblock
Persistent Volume (PV) is available:$ oc get pv | grep localblock
Example output:
NAME CAPACITY ACCESSMODES RECLAIMPOLICY STATUS CLAIM STORAGECLASS AGE local-pv-3e8964d3 500Gi RWO Delete Bound ocs-deviceset-localblock-2-data-0-mdbg9 localblock 25h local-pv-414755e0 500Gi RWO Delete Bound ocs-deviceset-localblock-1-data-0-4cslf localblock 25h local-pv-b481410 500Gi RWO Delete Available localblock 3m24s local-pv-5c9b8982 500Gi RWO Delete Bound ocs-deviceset-localblock-0-data-0-g2mmc localblock 25h
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required.
Identify the Persistent Volume Claim (PVC):
$ osd_id_to_remove=1
$ oc get -n openshift-storage -o yaml deployment rook-ceph-osd-${<osd_id_to_remove>} | grep ceph.rook.io/pvc
where,
<osd_id_to_remove>
is the integer in the pod name immediately after therook-ceph-osd
prefix.In this example, the deployment name is
rook-ceph-osd-1
.Example output:
ceph.rook.io/pvc: ocs-deviceset-localblock-0-data-0-g2mmc ceph.rook.io/pvc: ocs-deviceset-localblock-0-data-0-g2mmc
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix. You can add comma separated OSD IDs in the command to remove more than one OSD, for example,FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.WarningThis step results in the OSD being completely removed from the cluster. Ensure that the correct value of
osd_id_to_remove
is provided.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job has succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails, and the pod is not in the expectedCompleted
state, check the pod logs for further debugging.For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Delete the PV associated with the failed node.
Identify the PV associated with the PVC:
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-5c9b8982 500Gi RWO Delete Released openshift-storage/ocs-deviceset-localblock-0-data-0-g2mmc localblock 24h worker-0
The PVC name must be identical to the name that is obtained while removing the failed OSD from the cluster.
If there is a PV in
Released
state, delete it:# oc delete pv <persistent_volume>
For example:
# oc delete pv local-pv-5c9b8982
Example output:
persistentvolume "local-pv-5c9b8982" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in Running state:
-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in Running state.
Ensure that the new incremental
mon
is created and is in the Running state:$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-b-74f6dc9dd6-4llzq 1/1 Running 0 6h14m rook-ceph-mon-c-74948755c-h7wtx 1/1 Running 0 4h24m rook-ceph-mon-d-598f69869b-4bv49 1/1 Running 0 162m
The OSD and monitor pod might take several minutes to get to the
Running
state.Verify that the new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.4. Replacing storage nodes on VMware infrastructure
To replace an operational node, see:
To replace a failed node,see:
2.4.1. Replacing an operational node on VMware user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure, resources, and disks to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
Identify the node, and get the labels on the node that you need to replace:
$ oc get nodes --show-labels | grep <node_name>
<node_name>
- Specify the name of node that you need to replace.
Identify the monitor pod (if any), and OSDs that are running in the node that you need to replace:
$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon <node_name>
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
Delete the node:
$ oc delete node <node_name>
- Log in to VMware vSphere and terminate the Virtual Machine (VM) that you have identified.
- Create a new VM on VMware vSphere with the required infrastructure. See Infrastructure requirements.
- Create a new OpenShift Container Platform worker node using the new VM.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Identify the namespace where OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a new worker node to the
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumediscovery auto-discover-devices
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed, andnewnode.example.com
is the new node.Determine the
localVolumeSet
to edit:# oc get -n $local_storage_project localvolumeset
Example output:
NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumeset localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In the this example,
server3.example.com
is removed andnewnode.example.com
is the new node.
Verify that the new
localblock
Persistent Volume (PV) is available:$oc get pv | grep localblock | grep Available
Example output:
local-pv-551d950 512Gi RWO Delete Available localblock 26s
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix.You can add comma separated OSD IDs in the command to remove more than one OSD, for example,
FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails, and the pod is not in the expectedCompleted
state, check the pod logs for further debugging:For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Identify the Persistent Volume (PV) associated with the Persistent Volume Claim (PVC):
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-d6bf175b 1490Gi RWO Delete Released openshift-storage/ocs-deviceset-0-data-0-6c5pw localblock 2d22h compute-1
If there is a PV in
Released
state, delete it:# oc delete pv <persistent_volume>
For example:
# oc delete pv local-pv-d6bf175b
Example output:
persistentvolume "local-pv-d9c5cbd6" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in
Running
state:-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in
Running
state.Ensure that the new incremental
mon
is created, and is in theRunning
state:$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-a-cd575c89b-b6k66 2/2 Running 0 38m rook-ceph-mon-b-6776bc469b-tzzt8 2/2 Running 0 38m rook-ceph-mon-d-5ff5d488b5-7v8xh 2/2 Running 0 4m8s
OSD and monitor pod might take several minutes to get to the
Running
state.Verify that new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.4.2. Replacing an operational node on VMware installer-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with the similar infrastructure, resources, and disks to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the node that you need to replace. Take a note of its Machine Name.
Get labels on the node:
$ oc get nodes --show-labels | grep <node_name>
<node_name>
- Specify the name of node that you need to replace.
Identify the
mon
(if any), and Object Storage Devices (OSDs) that are running in the node:$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods that you identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon <node_name>
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
- Click Compute → Machines. Search for the required machine.
- Besides the required machine, click Action menu (⋮) → Delete Machine.
- Click Delete to confirm the machine deletion. A new machine is automatically created.
Wait for the new machine to start and transition into Running state.
ImportantThis activity might take at least 5 - 10 minutes or more.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
- Physically add a new device to the node.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Identify the namespace where the OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a new worker node to the
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node and remove the failed node.# oc edit -n $local_storage_project localvolumediscovery auto-discover-devices
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed, andnewnode.example.com
is the new node.Determine the
localVolumeSet
you need to edit:# oc get -n $local_storage_project localvolumeset
Example output:
NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node and remove the failed node:# oc edit -n $local_storage_project localvolumeset localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed, andnewnode.example.com
is the new node.
Verify that the new
localblock
Persistent Volume (PV) is available:$ oc get pv | grep localblock | grep Available
Example output:
local-pv-551d950 512Gi RWO Delete Available localblock 26s
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix.You can add comma separated OSD IDs in the command to remove more than one OSD, for example,
FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails and the pod is not in the expectedCompleted
state, check the pod logs for further debugging.For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Identify the PV associated with the Persistent Volume Claim (PVC):
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-d6bf175b 1490Gi RWO Delete Released openshift-storage/ocs-deviceset-0-data-0-6c5pw localblock 2d22h compute-1
If there is a PV in
Released
state, delete it:# oc delete pv <persistent_volume>
For example:
# oc delete pv local-pv-d6bf175b
Example output:
persistentvolume "local-pv-d9c5cbd6" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in
Running
state:-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in
Running
state.Ensure that the new incremental
mon
is created and is in theRunning
state.$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-a-cd575c89b-b6k66 2/2 Running 0 38m rook-ceph-mon-b-6776bc469b-tzzt8 2/2 Running 0 38m rook-ceph-mon-d-5ff5d488b5-7v8xh 2/2 Running 0 4m8s
OSD and monitor pod might take several minutes to get to the
Running
state.Verify that new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.4.3. Replacing a failed node on VMware user-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with similar infrastructure, resources, and disks to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
Identify the node, and get the labels on the node that you need to replace:
$ oc get nodes --show-labels | grep <node_name>
<node_name>
- Specify the name of node that you need to replace.
Identify the monitor pod (if any), and OSDs that are running in the node that you need to replace:
$ oc get pods -n openshift-storage -o wide | grep -i <node_name>
Scale down the deployments of the pods identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon <node_name>
Remove the pods which are in
Terminating
state:$ oc get pods -A -o wide | grep -i <node_name> | awk '{if ($4 == "Terminating") system ("oc -n " $1 " delete pods " $2 " --grace-period=0 " " --force ")}'
Drain the node:
$ oc adm drain <node_name> --force --delete-emptydir-data=true --ignore-daemonsets
Delete the node:
$ oc delete node <node_name>
- Log in to VMware vSphere and terminate the Virtual Machine (VM) that you have identified.
- Create a new VM on VMware vSphere with the required infrastructure. See Infrastructure requirements.
- Create a new OpenShift Container Platform worker node using the new VM.
Check for the Certificate Signing Requests (CSRs) related to OpenShift Container Platform that are in
Pending
state:$ oc get csr
Approve all the required OpenShift Container Platform CSRs for the new node:
$ oc adm certificate approve <certificate_name>
<certificate_name>
- Specify the name of the CSR.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node <new_node_name> cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Identify the namespace where OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a new worker node to the
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumediscovery auto-discover-devices
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed, andnewnode.example.com
is the new node.Determine the
localVolumeSet
to edit:# oc get -n $local_storage_project localvolumeset
Example output:
NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node, and remove the failed node:# oc edit -n $local_storage_project localvolumeset localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - newnode.example.com [...]
Remember to save before exiting the editor.
In the this example,
server3.example.com
is removed andnewnode.example.com
is the new node.
Verify that the new
localblock
Persistent Volume (PV) is available:$oc get pv | grep localblock | grep Available
Example output:
local-pv-551d950 512Gi RWO Delete Available localblock 26s
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix.You can add comma separated OSD IDs in the command to remove more than one OSD, for example,
FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails, and the pod is not in the expectedCompleted
state, check the pod logs for further debugging:For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Identify the Persistent Volume (PV) associated with the Persistent Volume Claim (PVC):
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-d6bf175b 1490Gi RWO Delete Released openshift-storage/ocs-deviceset-0-data-0-6c5pw localblock 2d22h compute-1
If there is a PV in
Released
state, delete it:# oc delete pv <persistent_volume>
For example:
# oc delete pv local-pv-d6bf175b
Example output:
persistentvolume "local-pv-d9c5cbd6" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=<failed_node_name> -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in
Running
state:-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in
Running
state.Ensure that the new incremental
mon
is created, and is in theRunning
state:$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-a-cd575c89b-b6k66 2/2 Running 0 38m rook-ceph-mon-b-6776bc469b-tzzt8 2/2 Running 0 38m rook-ceph-mon-d-5ff5d488b5-7v8xh 2/2 Running 0 4m8s
OSD and monitor pod might take several minutes to get to the
Running
state.Verify that new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.
2.4.4. Replacing a failed node on VMware installer-provisioned infrastructure
Prerequisites
- Ensure that the replacement nodes are configured with the similar infrastructure, resources, and disks to the node that you replace.
- You must be logged into the OpenShift Container Platform cluster.
Procedure
- Log in to the OpenShift Web Console, and click Compute → Nodes.
- Identify the node that you need to replace. Take a note of its Machine Name.
Get the labels on the node:
$ oc get nodes --show-labels | grep _<node_name>_
<node_name>
- Specify the name of node that you need to replace.
Identify the
mon
(if any) and Object Storage Devices (OSDs) that are running in the node:$ oc get pods -n openshift-storage -o wide | grep -i _<node_name>_
Scale down the deployments of the pods identified in the previous step:
For example:
$ oc scale deployment rook-ceph-mon-c --replicas=0 -n openshift-storage
$ oc scale deployment rook-ceph-osd-0 --replicas=0 -n openshift-storage
$ oc scale deployment --selector=app=rook-ceph-crashcollector,node_name=<node_name> --replicas=0 -n openshift-storage
Mark the node as unschedulable:
$ oc adm cordon _<node_name>_
Remove the pods which are in
Terminating
state:$ oc get pods -A -o wide | grep -i _<node_name>_ | awk '{if ($4 == "Terminating") system ("oc -n " $1 " delete pods " $2 " --grace-period=0 " " --force ")}'
Drain the node:
$ oc adm drain _<node_name>_ --force --delete-emptydir-data=true --ignore-daemonsets
- Click Compute → Machines. Search for the required machine.
- Besides the required machine, click Action menu (⋮) → Delete Machine.
- Click Delete to confirm the machine is deleted. A new machine is automatically created.
Wait for the new machine to start and transition into Running state.
ImportantThis activity might take at least 5 - 10 minutes or more.
- Click Compute → Nodes in the OpenShift Web Console. Confirm that the new node is in Ready state.
- Physically add a new device to the node.
Apply the OpenShift Data Foundation label to the new node using any one of the following:
- From the user interface
- For the new node, click Action Menu (⋮) → Edit Labels.
-
Add
cluster.ocs.openshift.io/openshift-storage
, and click Save.
- From the command-line interface
- Apply the OpenShift Data Foundation label to the new node:
$ oc label node _<new_node_name>_ cluster.ocs.openshift.io/openshift-storage=""
<new_node_name>
- Specify the name of the new node.
Identify the namespace where the OpenShift local storage operator is installed, and assign it to the
local_storage_project
variable:$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
For example:
$ local_storage_project=$(oc get csv --all-namespaces | awk '{print $1}' | grep local)
echo $local_storage_project
Example output:
openshift-local-storage
Add a new worker node to the
localVolumeDiscovery
andlocalVolumeSet
.Update the
localVolumeDiscovery
definition to include the new node and remove the failed node:# oc edit -n $local_storage_project localvolumediscovery auto-discover-devices
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - **newnode.example.com** [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed andnewnode.example.com
is the new node.Determine the
localVolumeSet
you need to edit.# oc get -n $local_storage_project localvolumeset
Example output:
NAME AGE localblock 25h
Update the
localVolumeSet
definition to include the new node and remove the failed node:# oc edit -n $local_storage_project localvolumeset localblock
Example output:
[...] nodeSelector: nodeSelectorTerms: - matchExpressions: - key: kubernetes.io/hostname operator: In values: - server1.example.com - server2.example.com #- server3.example.com - **newnode.example.com** [...]
Remember to save before exiting the editor.
In this example,
server3.example.com
is removed andnewnode.example.com
is the new node.
Verify that the new
localblock
PV is available:$ oc get pv | grep localblock | grep Available
Example output:
local-pv-551d950 512Gi RWO Delete Available localblock 26s
Navigate to the
openshift-storage
project:$ oc project openshift-storage
Remove the failed OSD from the cluster. You can specify multiple failed OSDs if required:
$ oc process -n openshift-storage ocs-osd-removal \ -p FAILED_OSD_IDS=<failed_osd_id> | oc create -f -
<failed_osd_id>
Is the integer in the pod name immediately after the
rook-ceph-osd
prefix.You can add comma separated OSD IDs in the command to remove more than one OSD, for example,
FAILED_OSD_IDS=0,1,2
.The
FORCE_OSD_REMOVAL
value must be changed totrue
in clusters that only have three OSDs, or clusters with insufficient space to restore all three replicas of the data after the OSD is removed.
Verify that the OSD was removed successfully by checking the status of the
ocs-osd-removal-job
pod.A status of
Completed
confirms that the OSD removal job succeeded.# oc get pod -l job-name=ocs-osd-removal-job -n openshift-storage
Ensure that the OSD removal is completed.
$ oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1 | egrep -i 'completed removal'
Example output:
2022-05-10 06:50:04.501511 I | cephosd: completed removal of OSD 0
ImportantIf the
ocs-osd-removal-job
fails and the pod is not in the expectedCompleted
state, check the pod logs for further debugging:For example:
# oc logs -l job-name=ocs-osd-removal-job -n openshift-storage --tail=-1
Identify the PV associated with the Persistent Volume Claim (PVC):
# oc get pv -L kubernetes.io/hostname | grep localblock | grep Released
Example output:
local-pv-d6bf175b 1490Gi RWO Delete Released openshift-storage/ocs-deviceset-0-data-0-6c5pw localblock 2d22h compute-1
If there is a PV in
Released
state, delete it:# oc delete pv _<persistent_volume>_
For example:
# oc delete pv local-pv-d6bf175b
Example output:
persistentvolume "local-pv-d9c5cbd6" deleted
Identify the
crashcollector
pod deployment:$ oc get deployment --selector=app=rook-ceph-crashcollector,node_name=_<failed_node_name>_ -n openshift-storage
If there is an existing
crashcollector
pod deployment, delete it:$ oc delete deployment --selector=app=rook-ceph-crashcollector,node_name=_<failed_node_name>_ -n openshift-storage
Delete the
ocs-osd-removal-job
:# oc delete -n openshift-storage job ocs-osd-removal-job
Example output:
job.batch "ocs-osd-removal-job" deleted
Verification steps
Verify that the new node is present in the output:
$ oc get nodes --show-labels | grep cluster.ocs.openshift.io/openshift-storage= |cut -d' ' -f1
Click Workloads → Pods. Confirm that at least the following pods on the new node are in
Running
state:-
csi-cephfsplugin-*
-
csi-rbdplugin-*
-
Verify that all other required OpenShift Data Foundation pods are in
Running
state.Ensure that the new incremental
mon
is created, and is in theRunning
state:$ oc get pod -n openshift-storage | grep mon
Example output:
rook-ceph-mon-a-cd575c89b-b6k66 2/2 Running 0 38m rook-ceph-mon-b-6776bc469b-tzzt8 2/2 Running 0 38m rook-ceph-mon-d-5ff5d488b5-7v8xh 2/2 Running 0 4m8s
OSD and monitor pod might take several minutes to get to the
Running
state.Verify that new OSD pods are running on the replacement node:
$ oc get pods -o wide -n openshift-storage| egrep -i <new_node_name> | egrep osd
Optional: If cluster-wide encryption is enabled on the cluster, verify that the new OSD devices are encrypted.
For each of the new nodes identified in the previous step, do the following:
Create a debug pod and open a chroot environment for the one or more selected hosts:
$ oc debug node/<node_name>
$ chroot /host
Display the list of available block devices:
$ lsblk
Check for the
crypt
keyword beside the one or moreocs-deviceset
names.
- If the verification steps fail, contact Red Hat Support.