[CEPH] Fixes for the OSD defrag cronjob

Fix a naming issue with the cronjob's binary, and schedule the cron
job to run every 15 minutes for the gates. Additonally check to
to ensure we are only running on block devices. Also update the
script to work with ceph-volume created devices.

Change-Id: I8aedab0ac41c191ef39a08034fff3278027d7520

ceph-osd/templates/bin/utils/_defragOSDs.sh.tpl

@@ -31,14 +31,15 @@ if [ "x${ARG}" == "xcron" ]; then
   done
 fi
 
-if [ "x${ARG}" == "xdefrag" ]; then
+if [ "x${ARG}" == "xdefrag" ] && [ "x${STORAGE_TYPE%-*}" == "xblock" ]; then
+  OSD_DEVICE=$(readlink -f ${STORAGE_LOCATION})
+  ODEV=$(echo ${OSD_DEVICE} | sed 's/[0-9]//g' | cut -f 3 -d '/')
   OSD_PATH=$(cat /proc/mounts | awk '/ceph-/{print $2}')
-  OSD_DEVICE=$(cat /proc/mounts | awk '/ceph-/{print $1}')
   OSD_STORE=$(cat ${OSD_PATH}/type)
+  DATA_PART=$(cat /proc/mounts | awk '/ceph-/{print $1}')
 
-  ODEV=$(echo "${OSD_DEVICE}" | sed 's/\(.*[^0-9]\)[0-9]*$/\1/' | awk -F'/' '{print $3}')
   ODEV_ROTATIONAL=$(cat /sys/block/${ODEV}/queue/rotational)
-  ODEV_SCHEDULER=$(cat /sys/block/${ODEV}/queue/scheduler)
+  ODEV_SCHEDULER=$(cat /sys/block/${ODEV}/queue/scheduler | tr -d '[]')
 
   # NOTE(supamatt): TODO implement bluestore defrag options once it's available upstream
   if [ "${ODEV_ROTATIONAL}" -eq "1" ] && [ "x${OSD_STORE}" == "xfilestore" ]; then

ceph-osd/templates/cronjob-defragosds.yaml

@@ -84,12 +84,12 @@ spec:
                     fieldPath: metadata.namespace
               - name: KUBECTL_PARAM
                 value: {{ tuple $envAll "ceph" "ceph-defragosd" | include "helm-toolkit.snippets.kubernetes_kubectl_params" | indent 10 }}
-            command: ["/tmp/utils-defragosds.sh"]
+            command: ["/tmp/utils-defragOSDs.sh"]
             args: ["cron"]
             volumeMounts:
               - name: ceph-osd-bin
-                mountPath: /tmp/utils-defragosds.sh
+                mountPath: /tmp/utils-defragOSDs.sh
-                subPath: utils-defragosds.sh
+                subPath: utils-defragOSDs.sh
                 readOnly: true
           restartPolicy: Never
           hostNetwork: true

ceph-osd/values.yaml

@@ -272,4 +272,4 @@ manifests:
   daemonset_osd: true
   job_image_repo_sync: true
   helm_tests: true
-  cronjob_defragosds: true
+  cronjob_defragosds: false

tools/deployment/multinode/030-ceph.sh

@@ -74,9 +74,22 @@ conf:
         journal:
           type: directory
           location: /var/lib/openstack-helm/ceph/osd/journal-one
-
+jobs:
+  ceph_defragosds:
+    # Execute every 15 minutes for gates
+    cron: "*/15 * * * *"
+    history:
+      # Number of successful job to keep
+      successJob: 1
+      # Number of failed job to keep
+      failJob: 1
+    concurrency:
+      # Skip new job if previous job still active
+      execPolicy: Forbid
+    startingDeadlineSecs: 60
 manifests:
   cronjob_checkPGs: true
+  cronjob_defragosds: true
 EOF
 
 for CHART in ceph-mon ceph-osd ceph-client ceph-provisioners; do

tools/deployment/osh-infra-logging/020-ceph.sh

@@ -166,15 +166,27 @@ conf:
         journal:
           type: directory
           location: /var/lib/openstack-helm/ceph/osd/journal-one
-
 pod:
   replicas:
     mds: 1
     mgr: 1
     rgw: 1
-
+jobs:
+  ceph_defragosds:
+    # Execute every 15 minutes for gates
+    cron: "*/15 * * * *"
+    history:
+      # Number of successful job to keep
+      successJob: 1
+      # Number of failed job to keep
+      failJob: 1
+    concurrency:
+      # Skip new job if previous job still active
+      execPolicy: Forbid
+    startingDeadlineSecs: 60
 manifests:
   cronjob_checkPGs: true
+  cronjob_defragosds: true
 EOF
 
 for CHART in ceph-mon ceph-osd ceph-client ceph-provisioners; do

tools/deployment/tenant-ceph/030-ceph.sh

@@ -57,8 +57,22 @@ deployment:
   cephfs_provisioner: true
   client_secrets: false
   rgw_keystone_user_and_endpoints: false
+jobs:
+  ceph_defragosds:
+    # Execute every 15 minutes for gates
+    cron: "*/15 * * * *"
+    history:
+      # Number of successful job to keep
+      successJob: 1
+      # Number of failed job to keep
+      failJob: 1
+    concurrency:
+      # Skip new job if previous job still active
+      execPolicy: Forbid
+    startingDeadlineSecs: 60
 manifests:
   deployment_mds: false
+  cronjob_defragosds: true
 bootstrap:
   enabled: true
 conf:

tools/deployment/tenant-ceph/040-tenant-ceph.sh

@@ -88,8 +88,22 @@ storageclass:
     admin_secret_namespace: tenant-ceph
 bootstrap:
   enabled: true
+jobs:
+  ceph_defragosds:
+    # Execute every 15 minutes for gates
+    cron: "*/15 * * * *"
+    history:
+      # Number of successful job to keep
+      successJob: 1
+      # Number of failed job to keep
+      failJob: 1
+    concurrency:
+      # Skip new job if previous job still active
+      execPolicy: Forbid
+    startingDeadlineSecs: 60
 manifests:
   deployment_mds: false
+  cronjob_defragosds: true
 ceph_mgr_modules_config:
   prometheus:
     server_port: 9284