Active IQ Unified Manager Discussions
Active IQ Unified Manager Discussions
Welcome,
In OnCommand Console I created policy and dataset which use storage services to mirror backups on the other filer. With backups everything go ok - they are visible on both sides: locally and remote. Problem is when I want to mount backup snapshots. Mounting snapshot from local proceeds well. Mounting backup snapshots from remote site ends with errors:
Mounting backup 52 of the dataset test (905) on the host 192.168.100.240 (403).
Setting preferLunClone=true since some of the secondary qtrees required for the backupmount operation are present in the base Snapshot copy of a backup relationship.
Starting backup mount operation on Host service 'onComand.xxx.xxx(327) for the dataset 905.
Operation ID 'MountBackup_9fdde536-00b8-4f46-8c88-1d73a1e6e0f6_2013-07-24_1454' returned from hostservice 'onComand.xxx.xxx'(327) for backup mount.
OnCommandHSVMware: hsMount1 62c8f00b43202d8431ef3877e0d22db3: Starting mount request
OnCommandHSVMware: hsMount1 62c8f00b43202d8431ef3877e0d22db3: Storage system error: Cannot create LUN on a read-only volume or qtree
OnCommandHSVMware: hsMount1 62c8f00b43202d8431ef3877e0d22db3: FLOW-11019: Failure in NaStorageCloneActionWrapper: Cannot create LUN on a read-only volume or qtree (errno=9108)
OnCommandHSVMware: hsMount1 62c8f00b43202d8431ef3877e0d22db3: FLOW-11010: Operation transitioning to abort due to prior failure.
OnCommandHSVMware: hsMount1 62c8f00b43202d8431ef3877e0d22db3: FLOW-11011: Operation aborted
Source volume : vsimA:/vol_2_240 (created manually, LUN resides directly without qtrees)
Destination volume: vsimB:/vol_2_240_3 (created and snapmirrored by storage services)
Software is up to date.
Given the LUN error messages it seems that you are using LUNs.
If you were using NFS, there are two help topics available that might assist you in this restore task.
They can be found under "Managing virtual inventory" -> "Mounting and unmounting backups in a VMware environment".
1. Setting the preferred network interface
2. Mounting backups on remote VMware ESX server (7-Mode only)
1) If you take a new on-demand backup and transfer it to secondary are you able to restore from the older backup that already failed to restore?
2) Which versions of UM, ONTAP, and Host Package are you running?
Hi kryan
kryan wrote:
1) If you take a new on-demand backup and transfer it to secondary are you able to restore from the older backup that already failed to restore?
I use iSCSI. I take backup on demand and I am able to mount from secondary older backups. The most recent one I can restore only when I broke snapmirror relantionship. Then Host Service mount snapshot as a LUN (like lun clone command - in that same volume). Older backups are mount beyond volume (like vol clone command which is supported with FlexClone license)
2) Which versions of UM, ONTAP, and Host Package are you running?
Both filers are simulators with DOT 8.1.3 (with FlexClone license)
Unified Manager - 5.2
Host Service - 1.3.0.1537
VMware plug-in for Host Service - 1.1.0.0
Hi Jaroslaw - have you checked the igroup settings and LUN mapping for the ISCSI iqn's on the destination sim?
You should be able to map the LUN and access it as read only. The ONcommand error above looks as though its trying to create a LUN clone, which I would expect to fail if the volume is snapmirrored as its read only.
Hi Martin,
igroup, iscsi initiator, and copy of LUN are automatically done by DFM (which is invoked by Host Service). But igroup are defined after LUN becomes available. You get the point which I wrote in previous message: "why for last backup DFM is trying to make "lun copy" instated of "vol copy" like for rest of snapshots". lun copy of course fails because it is called on destination volume which is snapmirrored thus read-only. DFM should use vol copy or automatically break snapmirror relationship and resync after unmounting.
It does seem strange that DFM is trying to make a LUN clone. Probably worth having a look a the DFM release notes maybe. I have always mounted the snaps manually at DR with no issues.
The behavior you describe sounds like SDW bug 461675 (SDW is bundled and used for these tasks) however it should be fixed in HP 1.3 based on what I have read.
I recommend that you open a support case with the following data attached:
1) DFMDC
2) ONTAPWindDC on the HP server for the host service logs
Also ensure you have issued Autosupports from the controllers covering the failure window
Once opened, please post the case #.
Kryan,
Where I can find webpage with this bug? I searched on now.netapp.com without success.
There is no public report for this bug, so you will not find the information.
OK,
Thanks everyone for help.
Funnily enough that was going to be my next suggestion!