ocs-disk-gather

Gather disk information from OCP4 nodes for OCS 4 local storage setup

• there are 3 versions of disk gatherer currently available:

  1. ocs-disk-gatherer.yaml Everyting is created in default namespace which needs no scc as default project does not apply those yet (this might change in future) and pulls the image from legacy and deprecated registry.access.redhat.com and hence requires no pull secret. This might change in future as well
  2. ocs-disk-gatherer-sha-no-secret.yaml same as 1. but uses sha instead of tag (no tag implies latest) which might be needed in some circomstances, but might soon be outdated, so mind security issues
  3. ocs-disk-gatherer-own-project.yaml This yaml creates
     • a namespace called ocs-disk-gatherer
     • a service account called ocs-disk-gatherer-sa
     • a scc called ocs-disk-gatherer-sa to be attached to the service account
     • a pull secret to fetch the RHEL 8 ubi image from the registry you need to get a pull secret from https://access.redhat.com/terms-based-registry/ and add it to the ubi-pull-secret section in the ocs-disk-gatherer.yaml

• all of the three choices from above create a daemon set called ocs-disk-gatherer that automatically deploys on every node labeld with cluster.ocs.openshift.io/openshift-storage:"" The ds will start a loop refreshing all 10 minutes the disks it found, output to look like this:

• check daemonset is running

# oc get ds -o wide
NAME                DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR                                 AGE   CONTAINERS   IMAGES                                SELECTOR
ocs-disk-gatherer   3         3         3       3            3           cluster.ocs.openshift.io/openshift-storage=   37s   collector    registry.access.redhat.com/ubi8/ubi   name=ocs-disk-gatherer
# 

• check pods running

# oc get po -o wide
NAME                      READY   STATUS    RESTARTS   AGE   IP            NODE                              NOMINATED NODE   READINESS GATES
ocs-disk-gatherer-4bjg5   1/1     Running   0          42s   10.131.2.12   cluster-jfb49-workerocs-0-plsbc   <none>           <none>
ocs-disk-gatherer-bhhlb   1/1     Running   0          42s   10.129.4.15   cluster-jfb49-workerocs-0-lnq2p   <none>           <none>
ocs-disk-gatherer-nwhhn   1/1     Running   0          42s   10.128.4.21   cluster-jfb49-workerocs-0-szp97   <none>           <none>
# 

• check logs on a certain pod

# oc logs ocs-disk-gatherer-bhhlb
          # NODE:cluster-jfb49-workerocs-0-lnq2p
          # sda : 120G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_143fdada-9a4e-47de-9
          # sdb : 25G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_c6c84893-9de0-402c-a
          # sdc : 1.2T
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_2bfc0ad1-812d-4702-8
 -------------------------------------------
# 

• check logs on all pods running at once

# oc logs -l name=ocs-disk-gatherer --tail=-1 --since=10m 
(or # oc logs -l name=ocs-disk-gatherer --tail=-1 --since=10m -n ocs-disk-gatherer)
          # NODE:cluster-jfb49-workerocs-0-lnq2p
          # sda : 120G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_143fdada-9a4e-47de-9
          # sdb : 25G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_c6c84893-9de0-402c-a
          # sdc : 1.2T
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_2bfc0ad1-812d-4702-8
 -------------------------------------------
          # NODE:cluster-jfb49-workerocs-0-szp97
          # sda : 120G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_26e6a4c3-5faf-4052-9
          # sdb : 1.2T
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_da04fbd6-dc2f-4a31-b
          # sdc : 25G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_1b5eeeae-f9a2-467f-9
 -------------------------------------------
          # NODE:cluster-jfb49-workerocs-0-plsbc
          # sda : 120G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_e5f05350-bddb-409e-a
          # sdb : 1.2T
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_afa53a80-a2a0-47d6-9
          # sdc : 25G
        - /dev/disk/by-id/scsi-0QEMU_QEMU_HARDDISK_e690e3a2-4bc0-4deb-b
 -------------------------------------------
#

• run one of the following to get the version of your choics:

  • wget https://raw.githubusercontent.com/dmoessne/ocs-disk-gather/master/ocs-disk-gatherer.yaml
  • wget https://raw.githubusercontent.com/dmoessne/ocs-disk-gather/master/ocs-disk-gatherer-sha-no-secret.yaml
  • wget https://raw.githubusercontent.com/dmoessne/ocs-disk-gather/master/ocs-disk-gatherer-own-project.yaml

  In case you've chosen the last one, don't forget to to modify the yaml and put your secret in :

# vim ocs-disk-gatherer-own-project.yaml
...
---
apiVersion: v1
kind: Secret
metadata:
  name: ubi-pull-secret
  namespace: ocs-disk-gatherer
data:
  .dockerconfigjson: <redacted, fetch from https://access.redhat.com/terms-based-registry/ >
type: kubernetes.io/dockerconfigjson
---
...

then run oc create -f < your donloaded file >

• You can leave the pods running and in case you add new disks the logs will be refreshed after about 10 minutes, however, it is not recommended at least for the versions running in default namespace.

• ds can be removed with oc delete -f < your donloaded file >