VPlex 5.3 : tips & tricks (CLI)

Pour terminer cette série de billets sur GeoSynchrony 5.3, voici une liste, que je mettrai à jour régulièrement, des nombreuses améliorations disponibles au sein de l’interface CLI de VPlex. Ces tips sont fournis dans le désordre (nous somme le 31 Décembre, et il est 16h00 … :) ).

configuration sync-time
J’ai reçu quelques instructions me demandant, dans la mesure du possible, de ne pas trop m’épancher sur cette fonction, mais bon, je n’ai pas pu résister tant elle manquait précédemment … on peut désormais synchroniser les horloges de nos clusters via NTP !!! Je vous laisse découvrir cette fabuleuse feature … et un petit clin d’oeil à Emmanuel pour le coup ;) . Ceci étant, les clusters étant tout de même très intriqués, je vous conseille de le demander directement à la hotline EMC, de manière à ne prendre aucun risque.

VPlexcli:/> configuration sync-time -h
synopsis: sync-time []

This command will synchronize the clock of the management server via NTP. On Cluster 2 only the public IP address of the Cluster 1 Management Server should be used.

options (* = required):
  -h | --help
          Displays the usage for this command.
  --verbose
          Provides more output during command execution.  This may not have any effect for some commands.
  -i | --remote-ip= 
          The IP address of the remote NTP server.
  -f | --force
          Skip verification prompt.

unclaim :
La fonction « unclaim », permettant de libérer un storage-volume de sa LUN de rattachement dispose désormais d’une option « –return-to-pool », qui permet, si ce storage-volume a été provisionné grâce au mécanisme automatique, de décommissionner aussi l’appartenance de la LUN au storage group VNX.

VPlexcli:/clusters/cluster-1/storage-elements/storage-volumes> unclaim -h
synopsis: unclaim [] 

Unclaims the specified previously claimed storage-volumes.

options (* = required):
  -h | --help
          Displays the usage for this command.
  --verbose
          Provides more output during command execution. This may not have any effect for some commands.
  -r | --return-to-pool
          For VIAS-based storage-volumes, this will return the storage capacity of each VIAS-based volume to the pool on the corresponding storage-array.
  --batch-size= 
          The maximum number of storage-volumes to unclaim at once.

positional arguments (* = required):
  -d | --storage-volumes=  [,  ...]
          The storage-volumes to be unclaimed.

storage-tool compose :
Une nouvelle directive permettant d’utiliser la fonction de provisioning direct des baies à travers l’interface CLI. A essayer dès que possible tant il semble que son potentiel soit très riches !

VPlexcli:/clusters/cluster-1/storage-elements/array-providers> storage-tool compose -h
synopsis: compose []   

Creates a virtual-volume atop the specified storage-volumes, building all intermediate extents, local and distributed devices as necessary.

options (* = required):
  -h | --help
          Displays the usage for this command.
  --verbose
          Provides more output during command execution. This may not have any effect for some commands.
  -m | --source-mirror= 
          Specifies the storage-volume to use as a source mirror when creating local and distributed devices.
  -c | --consistency-group= 
          Specifies a consistency-group to which the new virtual-volume should be added. The new virtual-volume's global geometry must be compatible with the consistency-group's storage-at-clusters
          attribute.
  -v | --storage-views=  [,  ...]
          Specifies the storage-views to which the new virtual-volume will be added. The new virtual-volume's global geometry must be compatible with the storage-views' locality.

positional arguments (* = required):
  *-n | --name= 
          Specifies the name to use for the new virtual-volume. Must be unique across the system.
  *-g | --geometry= 
          Specifies the desired geometry to use for the local-devices at each cluster. Must be one of 'raid-0', 'raid-1', or 'raid-c'.
  *-d | --storage-volumes=  [,  ...]
          Specifies the storage-volumes from which to build the virtual-volume. These may be claimed but must be unused.

For raid-1 local-devices a maximum of eight legs may be specified.

In the event of an error the command will not attempt to perform a roll-back and destroy any intermediate storage objects it has created. If cleanup is necessary, use the 'show-use-hierarchy'
command on each storage-volume to identify all residual objects and delete each one manually.

If specified, --source-mirror will be used as a source mirror when creating local AND distributed raid-1 devices. This will trigger a rebuild from the source mirror to all other mirrors of the
raid-1 device (local and distributed). While the rebuild is in progress the new virtual-volume (and supporting local and/or distributed devices) will be in a degraded state; this is normal. This
option only applies to raid-1 devices (local or distributed). The --source-mirror may also appear in --storage-volumes.

Note that if the new virtual-volume's global geometry is not compatible with the specified consistency-group or storage-views the virtual-volume will not be created. However, failure to add the new
virtual-volume to the specified consistency-group or storage-views does not constitute an overall failure to create the storage and will not be reported as such.