Subscribe

Could not get primary host stab by ip/name in PM?

Hi,

I have NFS volume called sap_ZZZ_redorepository (mounted from some box), i'm trying to create a SnapVault relationship but by using PM. So i go to the PM, create a new DataSet named redo_XXX_ZZZ, later assign PM policy to it (i want it to be SnapVaulted daily)

Then the'Create relationship' job is succededed but has warning "redo_XXX_ZZZ:Could not get primary host stab by ip/name ControllerNameHere".

DataOnTap 7.3.2, PM 3.8.1

From dfmserver.log:

Aug 31 12:31:36 [dfmserver: INFO]: Thread 0x1574: Created storage set redo_XXX_ZZZ:root (13650).
Aug 31 12:31:36 [dfmserver: INFO]: Thread 0x1574: Created dataset redo_XXX_ZZZ (13651).
Aug 31 12:31:48 [dfmserver: INFO]: Thread 0x1fd0: Modified dataset redo_XXX_ZZZ(13651): Changed description to JK2.
Aug 31 12:31:48 [dfmserver: INFO]: Thread 0x1fd0: Modified dataset redo_XXX_ZZZ(13651): Changed owner to Jakub.
Aug 31 12:31:48 [dfmserver: INFO]: Thread 0x1fd0: Modified dataset redo_XXX_ZZZ(13651): Changed contact to Jakub.
Aug 31 12:31:48 [dfmserver: INFO]: Thread 0x1fd0: Added Volume ControllerNameHere:/sap_ZZZ_redorepository(13092) to storage set redo_XXX_ZZZ:root (13650).
Aug 31 12:31:48 [dfmserver: INFO]: Thread 0x1fd0: Modified export settings for node redo_XXX_ZZZ:root of dataset redo_XXX_ZZZ (13650).
Aug 31 12:32:17 [dfmserver: INFO]: Thread 0x6ac: Created storage set redo_XXX_ZZZ:Backup (13652).
Aug 31 12:32:17 [dfmserver: INFO]: Thread 0x6ac: Set data protection policy for dataset 'redo_XXX_ZZZ' to 'ABC_Policy_LAB_JK2'.
Aug 31 12:32:17 [dfmserver: INFO]: Thread 0x6ac: Set node 'Backup' of data protection policy 'ABC_Policy_LAB_JK2' for dataset 'redo_XXX_ZZZ' to 'redo_XXX_ZZZ:Backup'.
Aug 31 12:32:17 [dfmserver: INFO]: Thread 0x6ac: Modified export settings for node redo_XXX_ZZZ:Backup of dataset redo_XXX_ZZZ (13652).
Aug 31 12:32:17 [dfmserver: INFO]: Thread 0x6ac: Added resource pool ResPool_Vault (3672) to storage set redo_XXX_ZZZ:Backup (13652).

From conformance.log:

Aug 31 12:32:17 [dfmserver: INFO]: Thread 0x16cc: Provision a new flexible volume (Actual run) of 660 GB for backup secondary node 'redo_XXX_ZZZ:Backup' (13652).
Aug 31 12:32:19 [dfmserver: INFO]: Thread 0x16cc: Inserted newly provisioned flexible volume 'redo_XXX_ZZZ_backup' (13653).
Aug 31 12:32:25 [dfmserver:ERROR]: Thread 0x16cc: Command 'exportfs -q /vol/redo_XXX_ZZZ_backup' failed. Reason: timed out

The command itself (exportfs -q <vol>) does not return anything on the controller.The volume exists:

ControllerNameHere> exportfs -q /vol/redo_XXX_ZZZ_backup
ControllerNameHere> df -g /vol/redo_XXX_ZZZ_jk2_backup
Filesystem               total       used      avail capacity  Mounted on
/vol/redo_XXX_ZZZ_backup/      660GB        0GB      659GB       0%  /vol/redo_XXX_ZZZ_backup/
/vol/redo_XXX_ZZZ_backup/.snapshot        0GB        0GB        0GB     ---%  /vol/redo_XXX_ZZZ_backup/.snapshot
ControllerNameHere>

SnapVault relationship also exists

ControllerNameHere:/vol/sap_jk2_redorepository/-                kopdslab04d:/vol/redo_XXX_ZZZ_backup/redo_XXX_ZZZ_backup_ControllerNameHere_sap_ZZZ_redorep

It is snapvaulted, idle, snapvault snap schedule returns some awkward schedule.

I'm able to ping both controllers from PM Windows box. Protection for dataset is displaying 'Job Failure' and details are :

"The most recent protection job did not succeed. This status might also be
displayed if the baseline transfer could not be registered as a backup
version because either you are running Data ONTAP 7.2.3 or earlier, or there
was a short network outage after the baseline transfer was completed. In either
case, the next regularly scheduled or on-demand backup will be registered."

/etc/messages on controller has:

Tue Aug 31 16:32:18 UTC [ControllerNameHere: export.auto.update.disabled:warning]: /etc/exports was not updated for redo_XXX_ZZZ_backup when the vol create command was run. Please either manually update /etc/exports or copy /etc/exports.new to it.

Indeed the .new file has NFS export line for redo_XXX_ZZZ_backup - the old one is not having this line. For me it looks like some bug/misconfig because PM is not able to refresh NFS exports after creating protected DataSet that has only this single volume (???) Why it is trying to exportfs it at all? The policy only has 'Backup' node (no Mirror).

Someone has clue? This timeout is mysterious.

-Jakub.

Re: Could not get primary host stab by ip/name in PM?

Hi

Can you check what’s the value for this options in the filer where it tried to create NFS exports.

nfs.export.auto-update

Also can you get us the details like

Dfpm dataset list –x for the error you mentioned.

Regards

adai

Re: Could not get primary host stab by ip/name in PM?

nfs.export.auto-update is off. If i understand if it would be "on" then every "vol create" would create a line in the /etc/exports . This is not what we want, because we have many vols dedicated only for FC.

dfpm output:

Id:                              13651
Name:                            redo_XXX_ZZZ
Policy:                          GSK_Policy_LAB_ZZZ
Description:                     ZZZ
Owner:                           Jakub
Contact:                         Jakub
Volume Qtree Name Prefix:
DR Capable:                      No
Requires Non Disruptive Restore: No

Node details:

   Node Name:           Primary data
   Resource Pools:
   Provisioning Policy:
   Time Zone:
   DR Capable:          No
   vFiler:

   Node Name:           Backup
   Resource Pools:      ABC_Vault
   Provisioning Policy:
   Time Zone:
   DR Capable:          No
   vFiler:

EDIT: what is more interesting that the schedule (assigned to the "Backup" node in the policy) was able to run Job of type "Remote backup", with all steps successfull, on the storage (on the destination volume) there are visible additional snapshots created by dfpm,

BTW: could you ellaborate a little bit, what exactly PM is doing when doing a "Remote backup", why it is doing "Snapshot create" (is it just snap create? with name of dfpm_base(..)), then SnapVault Start (what it does exactly, transfer of single?/all? snapshots from pri. volume?

Message was edited by: jakub.wartak