Subscribe

Troubleshoot performance issues Snapshot to blame?

Hello Everyone,

I have a FAS8040 that is presenting performance issues. I might be runing into some sort of misconfiguration issue. 

I can't find the snapshot policy for a particular volume (which is used as a DP Target)... Bear with me...

 

CLUSTER_NAME::> volume snapshot policy show -vserver SVMNAME
There are no entries matching your query.

CLUSTER_NAME::> volume snapshot show -vserver SVMNAME
---Blocks---
Vserver Volume Snapshot Size Total% Used%
-------- -------- ------------------------------------- -------- ------ -----
SVMNAME
SVMNAME_root
hourly.2016-01-19_2230 176KB 0% 33%
hourly.2016-01-20_1230 176KB 0% 33%
hourly.2016-01-20_2230 180KB 0% 33%
......
VOLUME_NAME_01
hourly.2014-09-16_2230 128KB 0% 0%
hourly.2014-09-17_1230 240KB 0% 0%
Snapshot_Daily.2015-11-28_0000 466.9MB 0% 0%
Snapshot_Daily.2015-12-04_0000 1.37GB 0% 0%%
Snapshot_Hourly.2016-01-26_0300 127.2MB 0% 0%
...
Vserver Volume Snapshot Size Total% Used%
-------- -------- ------------------------------------- -------- ------ -----
SVMNAME
DR_FS_APPFILES_01
Snapshot_Hourly.2016-01-26_0400 40.12MB 0% 0%
Snapshot_Hourly.2016-01-26_0500 24.45MB 0% 0%
...
snapmirror.cd78346d-3eaa-11e4-8be4-123478563412_2147484716.2016-01-26_144500
19.02MB 0% 0%
Snapshot_Hourly.2016-01-26_1500 209.5MB 0% 0%
snapmirror.cd78346d-3eaa-11e4-8be4-123478563412_2147484716.2016-01-26_154500
0B 0% 0%
126 entries were displayed.

 

It looks like I'm taking snapshots from this volume, althouh this one if for DR purposes. Here is what I mean:

 

Site 1, volume A gets replicated to Site B Volume DR_A (this one I'm looking for)

Then, it seems that I'm taking snapshots on Volume DR_A hourly and daily and what-not.... which doesn't seem to make much sense.... 

 

Do you agree? Can you help me?

There is any command that tells me WHAT/WHO created a snapshot, so I can figure out what is creating those snapshots.

 

Any help is appreciated. Thanks.

Re: Troubleshoot performance issues Snapshot to blame?

When you mirror a volume, all the snapshots on the source are copied over to the destination. The snaphots you are seeing are likely replicated from the source.

Re: Troubleshoot performance issues Snapshot to blame?

Hum! Didn't knew that. Thanks!

 

 I thought only that snapmirror snapshot would come... 

In any case, when I list the snapshot policy for the SVM, I can't see it. 

 

 

CLUSTERNAME::> volume snapshot policy show -vserver SNM_NAME
There are no entries matching your query.

 

But then again, 

 

CLUSTERNAME::> vserver show -vserver SVM_NAME -instance

Vserver: SVM_NAME
Vserver Type: data
...

Snapshot Policy: default
...

Allowed Protocols: cifs

CLUSTER_NAME::> snapshot policy show -policy default -instance

Vserver: CLUSTER_NAME
Snapshot Policy Name: default
Snapshot Policy Enabled: true
Policy Owner: cluster-admin
Comment: Default policy with hourly, daily & weekly schedules.
Total Number of Schedules: 1
Schedule Count Prefix SnapMirror Label
---------------------- ----- --------------------- -------------------
hourly 14 hourly hourly

 

 

But when I list the snapshots, I get a lot of "Daily" ones, and I can't figure out "what" is creating them. 

 

That's why I asked: "Is there any command that tells me WHAT/WHO created a particular snapshot??"

Re: Troubleshoot performance issues Snapshot to blame?

Just for clarification:

 

Volume A - the source volume - does this have the snapshots you are trying to identify? If it has the default snapshot schedule assigned, Data OnTap will be creating the snapshots.

Volume DR_A - the destination volume - this should show the same snapshots on the source volume if it is a regular mirror, or it could have additioanl ones if it is a vault mirror.

 

The snapmirror relationship will also generate snapshots, that relationship can be created via Unified Manager or the CLI. You can compare the snapmirror-label to see if it matches the snapshots you are seeing.

 

Have you shown the volume show for both the source and destination volumes?