Chapter 5. Gathering data about your cluster

Procedure

1. Obtain a list of cluster nodes:

   $ oc get nodes

   Copy to Clipboard Toggle word wrap

2. Enter into a debug session on the target node. This step instantiates a debug pod called <node_name>-debug:

   $ oc debug node/my-cluster-node

   Copy to Clipboard Toggle word wrap

   To enter into a debug session on the target node that is tainted with the NoExecute effect, add a toleration to a dummy namespace, and start the debug pod in the dummy namespace:

   $ oc new-project dummy

   Copy to Clipboard Toggle word wrap

   $ oc patch namespace dummy --type=merge -p '{"metadata": {"annotations": { "scheduler.alpha.kubernetes.io/defaultTolerations": "[{\"operator\": \"Exists\"}]"}}}'

   Copy to Clipboard Toggle word wrap

   $ oc debug node/my-cluster-node

   Copy to Clipboard Toggle word wrap

3. Set /host as the root directory within the debug shell. The debug pod mounts the host’s root file system in /host within the pod. By changing the root directory to /host, you can run binaries contained in the host’s executable paths:

   # chroot /host

   Copy to Clipboard Toggle word wrap
   Note

   OpenShift Container Platform 4.10 cluster nodes running Red Hat Enterprise Linux CoreOS (RHCOS) are immutable and rely on Operators to apply cluster changes. Accessing cluster nodes by using SSH is not recommended. However, if the OpenShift Container Platform API is not available, or the kubelet is not properly functioning on the target node, oc operations will be impacted. In such situations, it is possible to access nodes using ssh core@<node>.<cluster_name>.<base_domain> instead.

4. Start a toolbox container, which includes the required binaries and plugins to run sosreport:

   # toolbox

   Copy to Clipboard Toggle word wrap
   Note

   If an existing toolbox pod is already running, the toolbox command outputs 'toolbox-' already exists. Trying to start…​. Remove the running toolbox container with podman rm toolbox- and spawn a new toolbox container, to avoid issues with sosreport plugins.

5. Collect a sosreport archive.

   1. Run the sosreport command and enable the crio.all and crio.logs CRI-O container engine sosreport plugins:

      # sosreport -k crio.all=on -k crio.logs=on 

      1

      Copy to Clipboard Toggle word wrap

      1

      -k enables you to define sosreport plugin parameters outside of the defaults.

   2. Press Enter when prompted, to continue.
   3. Provide the Red Hat Support case ID. sosreport adds the ID to the archive’s file name.

   4. The sosreport output provides the archive’s location and checksum. The following sample output references support case ID 01234567:

      Your sosreport has been generated and saved in:
        /host/var/tmp/sosreport-my-cluster-node-01234567-2020-05-28-eyjknxt.tar.xz 

      1

      
      
      The checksum is: 382ffc167510fd71b4f12a4f40b97a4e

      Copy to Clipboard Toggle word wrap

      1

      The sosreport archive’s file path is outside of the chroot environment because the toolbox container mounts the host’s root directory at /host.

6. Provide the sosreport archive to Red Hat Support for analysis, using one of the following methods.

   • Upload the file to an existing Red Hat support case directly from an OpenShift Container Platform cluster.

     1. From within the toolbox container, run redhat-support-tool to attach the archive directly to an existing Red Hat support case. This example uses support case ID 01234567:

        # redhat-support-tool addattachment -c 01234567 /host/var/tmp/my-sosreport.tar.xz 

        1

        Copy to Clipboard Toggle word wrap

        1

        The toolbox container mounts the host’s root directory at /host. Reference the absolute path from the toolbox container’s root directory, including /host/, when specifying files to upload through the redhat-support-tool command.

   • Upload the file to an existing Red Hat support case.

     1. Concatenate the sosreport archive by running the oc debug node/<node_name> command and redirect the output to a file. This command assumes you have exited the previous oc debug session:

        $ oc debug node/my-cluster-node -- bash -c 'cat /host/var/tmp/sosreport-my-cluster-node-01234567-2020-05-28-eyjknxt.tar.xz' > /tmp/sosreport-my-cluster-node-01234567-2020-05-28-eyjknxt.tar.xz 

        1

        Copy to Clipboard Toggle word wrap

        1

        The debug container mounts the host’s root directory at /host. Reference the absolute path from the debug container’s root directory, including /host, when specifying target files for concatenation.

        Note

        OpenShift Container Platform 4.10 cluster nodes running Red Hat Enterprise Linux CoreOS (RHCOS) are immutable and rely on Operators to apply cluster changes. Transferring a sosreport archive from a cluster node by using scp is not recommended. However, if the OpenShift Container Platform API is not available, or the kubelet is not properly functioning on the target node, oc operations will be impacted. In such situations, it is possible to copy a sosreport archive from a node by running scp core@<node>.<cluster_name>.<base_domain>:<file_path> <local_path>.

     2. Navigate to an existing support case within https://access.redhat.com/support/cases/.
     3. Select Attach files and follow the prompts to upload the file.

Procedure

1. Query bootkube.service journald unit logs from a bootstrap node during OpenShift Container Platform installation. Replace <bootstrap_fqdn> with the bootstrap node’s fully qualified domain name:

   $ ssh core@<bootstrap_fqdn> journalctl -b -f -u bootkube.service

   Copy to Clipboard Toggle word wrap
   Note

   The bootkube.service log on the bootstrap node outputs etcd connection refused errors, indicating that the bootstrap server is unable to connect to etcd on control plane nodes. After etcd has started on each control plane node and the nodes have joined the cluster, the errors should stop.

2. Collect logs from the bootstrap node containers using podman on the bootstrap node. Replace <bootstrap_fqdn> with the bootstrap node’s fully qualified domain name:

   $ ssh core@<bootstrap_fqdn> 'for pod in $(sudo podman ps -a -q); do sudo podman logs $pod; done'

   Copy to Clipboard Toggle word wrap

Procedure

1. Query kubelet journald unit logs from OpenShift Container Platform cluster nodes. The following example queries control plane nodes only:

   $ oc adm node-logs --role=master -u kubelet  

   1

   Copy to Clipboard Toggle word wrap

   1

   Replace kubelet as appropriate to query other unit logs.

2. Collect logs from specific subdirectories under /var/log/ on cluster nodes.

   1. Retrieve a list of logs contained within a /var/log/ subdirectory. The following example lists files in /var/log/openshift-apiserver/ on all control plane nodes:

      $ oc adm node-logs --role=master --path=openshift-apiserver

      Copy to Clipboard Toggle word wrap

   2. Inspect a specific log within a /var/log/ subdirectory. The following example outputs /var/log/openshift-apiserver/audit.log contents from all control plane nodes:

      $ oc adm node-logs --role=master --path=openshift-apiserver/audit.log

      Copy to Clipboard Toggle word wrap

   3. If the API is not functional, review the logs on each node using SSH instead. The following example tails /var/log/openshift-apiserver/audit.log:

      $ ssh core@<master-node>.<cluster_name>.<base_domain> sudo tail -f /var/log/openshift-apiserver/audit.log

      Copy to Clipboard Toggle word wrap
      Note

      OpenShift Container Platform 4.10 cluster nodes running Red Hat Enterprise Linux CoreOS (RHCOS) are immutable and rely on Operators to apply cluster changes. Accessing cluster nodes by using SSH is not recommended. Before attempting to collect diagnostic data over SSH, review whether the data collected by running oc adm must gather and other oc commands is sufficient instead. However, if the OpenShift Container Platform API is not available, or the kubelet is not properly functioning on the target node, oc operations will be impacted. In such situations, it is possible to access nodes using ssh core@<node>.<cluster_name>.<base_domain>.

Procedure

1. Run a packet capture from the host network on some nodes by running the following command:

   $ oc adm must-gather \
       --dest-dir /tmp/captures \  <.>
       --source-dir '/tmp/tcpdump/' \  <.>
       --image registry.redhat.io/openshift4/network-tools-rhel8:latest \  <.>
       --node-selector 'node-role.kubernetes.io/worker' \  <.>
       --host-network=true \  <.>
       --timeout 30s \  <.>
       -- \
       tcpdump -i any \  <.>
       -w /tmp/tcpdump/%Y-%m-%dT%H:%M:%S.pcap -W 1 -G 300

   Copy to Clipboard Toggle word wrap

   <.> The --dest-dir argument specifies that oc adm must-gather stores the packet captures in directories that are relative to /tmp/captures on the client machine. You can specify any writable directory. <.> When tcpdump is run in the debug pod that oc adm must-gather starts, the --source-dir argument specifies that the packet captures are temporarily stored in the /tmp/tcpdump directory on the pod. <.> The --image argument specifies a container image that includes the tcpdump command. <.> The --node-selector argument and example value specifies to perform the packet captures on the worker nodes. As an alternative, you can specify the --node-name argument instead to run the packet capture on a single node. If you omit both the --node-selector and the --node-name argument, the packet captures are performed on all nodes. <.> The --host-network=true argument is required so that the packet captures are performed on the network interfaces of the node. <.> The --timeout argument and value specify to run the debug pod for 30 seconds. If you do not specify the --timeout argument and a duration, the debug pod runs for 10 minutes. <.> The -i any argument for the tcpdump command specifies to capture packets on all network interfaces. As an alternative, you can specify a network interface name.

2. Perform the action, such as accessing a web application, that triggers the network communication issue while the network trace captures packets.

3. Review the packet capture files that oc adm must-gather transferred from the pods to your client machine:

   tmp/captures
   ├── event-filter.html
   ├── ip-10-0-192-217-ec2-internal  

   1

   
   │   └── registry-redhat-io-openshift4-network-tools-rhel8-sha256-bca...
   │       └── 2022-01-13T19:31:31.pcap
   ├── ip-10-0-201-178-ec2-internal  

   2

   
   │   └── registry-redhat-io-openshift4-network-tools-rhel8-sha256-bca...
   │       └── 2022-01-13T19:31:30.pcap
   ├── ip-...
   └── timestamp

   Copy to Clipboard Toggle word wrap

   1 2

   The packet captures are stored in directories that identify the hostname, container, and file name. If you did not specify the --node-selector argument, then the directory level for the hostname is not present.

Procedure

1. Obtain a list of cluster nodes:

   $ oc get nodes

   Copy to Clipboard Toggle word wrap

2. Enter into a debug session on the target node. This step instantiates a debug pod called <node_name>-debug:

   $ oc debug node/my-cluster-node

   Copy to Clipboard Toggle word wrap

3. Set /host as the root directory within the debug shell. The debug pod mounts the host’s root file system in /host within the pod. By changing the root directory to /host, you can run binaries contained in the host’s executable paths:

   # chroot /host

   Copy to Clipboard Toggle word wrap
   Note

   OpenShift Container Platform 4.10 cluster nodes running Red Hat Enterprise Linux CoreOS (RHCOS) are immutable and rely on Operators to apply cluster changes. Accessing cluster nodes by using SSH is not recommended. However, if the OpenShift Container Platform API is not available, or the kubelet is not properly functioning on the target node, oc operations will be impacted. In such situations, it is possible to access nodes using ssh core@<node>.<cluster_name>.<base_domain> instead.

4. From within the chroot environment console, obtain the node’s interface names:

   # ip ad

   Copy to Clipboard Toggle word wrap

5. Start a toolbox container, which includes the required binaries and plugins to run sosreport:

   # toolbox

   Copy to Clipboard Toggle word wrap
   Note

   If an existing toolbox pod is already running, the toolbox command outputs 'toolbox-' already exists. Trying to start…​. To avoid tcpdump issues, remove the running toolbox container with podman rm toolbox- and spawn a new toolbox container.

6. Initiate a tcpdump session on the cluster node and redirect output to a capture file. This example uses ens5 as the interface name:

   $ tcpdump -nn -s 0 -i ens5 -w /host/var/tmp/my-cluster-node_$(date +%d_%m_%Y-%H_%M_%S-%Z).pcap  

   1

   Copy to Clipboard Toggle word wrap

   1

   The tcpdump capture file’s path is outside of the chroot environment because the toolbox container mounts the host’s root directory at /host.

7. If a tcpdump capture is required for a specific container on the node, follow these steps.

   1. Determine the target container ID. The chroot host command precedes the crictl command in this step because the toolbox container mounts the host’s root directory at /host:

      # chroot /host crictl ps

      Copy to Clipboard Toggle word wrap

   2. Determine the container’s process ID. In this example, the container ID is a7fe32346b120:

      # chroot /host crictl inspect --output yaml a7fe32346b120 | grep 'pid' | awk '{print $2}'

      Copy to Clipboard Toggle word wrap

   3. Initiate a tcpdump session on the container and redirect output to a capture file. This example uses 49628 as the container’s process ID and ens5 as the interface name. The nsenter command enters the namespace of a target process and runs a command in its namespace. because the target process in this example is a container’s process ID, the tcpdump command is run in the container’s namespace from the host:

      # nsenter -n -t 49628 -- tcpdump -nn -i ens5 -w /host/var/tmp/my-cluster-node-my-container_$(date +%d_%m_%Y-%H_%M_%S-%Z).pcap.pcap  

      1

      Copy to Clipboard Toggle word wrap

      1

      The tcpdump capture file’s path is outside of the chroot environment because the toolbox container mounts the host’s root directory at /host.

8. Provide the tcpdump capture file to Red Hat Support for analysis, using one of the following methods.

   • Upload the file to an existing Red Hat support case directly from an OpenShift Container Platform cluster.

     1. From within the toolbox container, run redhat-support-tool to attach the file directly to an existing Red Hat Support case. This example uses support case ID 01234567:

        # redhat-support-tool addattachment -c 01234567 /host/var/tmp/my-tcpdump-capture-file.pcap 

        1

        Copy to Clipboard Toggle word wrap

        1

        The toolbox container mounts the host’s root directory at /host. Reference the absolute path from the toolbox container’s root directory, including /host/, when specifying files to upload through the redhat-support-tool command.

   • Upload the file to an existing Red Hat support case.

     1. Concatenate the sosreport archive by running the oc debug node/<node_name> command and redirect the output to a file. This command assumes you have exited the previous oc debug session:

        $ oc debug node/my-cluster-node -- bash -c 'cat /host/var/tmp/my-tcpdump-capture-file.pcap' > /tmp/my-tcpdump-capture-file.pcap 

        1

        Copy to Clipboard Toggle word wrap

        1

        The debug container mounts the host’s root directory at /host. Reference the absolute path from the debug container’s root directory, including /host, when specifying target files for concatenation.

        Note

        OpenShift Container Platform 4.10 cluster nodes running Red Hat Enterprise Linux CoreOS (RHCOS) are immutable and rely on Operators to apply cluster changes. Transferring a tcpdump capture file from a cluster node by using scp is not recommended. However, if the OpenShift Container Platform API is not available, or the kubelet is not properly functioning on the target node, oc operations will be impacted. In such situations, it is possible to copy a tcpdump capture file from a node by running scp core@<node>.<cluster_name>.<base_domain>:<file_path> <local_path>.

     2. Navigate to an existing support case within https://access.redhat.com/support/cases/.
     3. Select Attach files and follow the prompts to upload the file.

Procedure

1. Set /host as the root directory within the debug shell. The debug pod mounts the host’s root file system in /host within the pod. By changing the root directory to /host, you can run binaries contained in the host’s executable paths:

   # chroot /host

   Copy to Clipboard Toggle word wrap

2. Start the toolbox container:

   # toolbox

   Copy to Clipboard Toggle word wrap

3. Install the additional package, such as wget:

   # dnf install -y <package_name>

   Copy to Clipboard Toggle word wrap

Procedure

1. Set /host as the root directory within the debug shell. The debug pod mounts the host’s root file system in /host within the pod. By changing the root directory to /host, you can run binaries contained in the host’s executable paths:

   # chroot /host

   Copy to Clipboard Toggle word wrap

2. Create a .toolboxrc file in the home directory for the root user ID:

   # vi ~/.toolboxrc

   Copy to Clipboard Toggle word wrap

   REGISTRY=quay.io                

   1

   
   IMAGE=fedora/fedora:33-x86_64   

   2

   
   TOOLBOX_NAME=toolbox-fedora-33  

   3

   Copy to Clipboard Toggle word wrap

   1

   Optional: Specify an alternative container registry.

   2

   Specify an alternative image to start.

   3

   Optional: Specify an alternative name for the toolbox container.

3. Start a toolbox container with the alternative image:

   # toolbox

   Copy to Clipboard Toggle word wrap
   Note

   If an existing toolbox pod is already running, the toolbox command outputs 'toolbox-' already exists. Trying to start…​. Remove the running toolbox container with podman rm toolbox- and spawn a new toolbox container, to avoid issues with sosreport plugins.