Subscribe

Unable to collect volume stats

Hi all,

I recently add a filer to the DFM. I have 4 volumes on the filer. But I was only able to collect the volume performance stats (e.g. Ops/sec, latency, data throughput....) for three volumes out of four. All other information like size, growth, snapshot are fine for all the 4 volumes. Is there a way that I can fix it so that I can collect the performance stats for the remaining one?

Thanks,

Jason

Re: Unable to collect volume stats

Performance Advisor collects statistics for all volumes. Hence it should

collect data for all 4 volumes. Was this volume added recently?

Performance data for any new volume should start showing up within 15

minutes.

Can you run "dfm perf data retrieve -o <volume-id> -C

volume:avg_latency" (I am assuming you are using DFM 4.0) on this fourth

volume and send us the output? If you can also run this command for the

other three volumes and "dfm volume list <filer-id>" and send us the

output, that would be helpful in analyzing the issue.

Regards

Harish

Re: Unable to collect volume stats

The volume was not created recently.

The DFM version is 3.7.1 and the command "dfm perf data retrieve" is invalid.

The for volume list, I was able to list all the four volume:

ID   Volume Name                          Type         Deleted  Deleted By
---- ------------------------------------ ------------ -------- ---------------
765 <hostname>:<volume 1>     Flexible     No
878 <hostname>:<volume 2>    Flexible     No
763 <hostname>:<volume 3>     Flexible     No
761 <hostname>:/vol0                Flexible     No

Re: Unable to collect volume stats

Since you did not mention the DFM version you were using, I assumed that you are using the latest 4.0 version.

Can you please follow these steps to help debug the issue:

1. Create a file called testpoints.txt in log directory (/opt/NTAPdfm/log on linux) with the following content:

testpoint -m pmhandlers -n xml-trace -e 1 -r 1

2. Wait for say 10 minutes.

3. Send us the contents of xml folder that got created in log directory (/opt/NTAPdfm/log/xml on linux)

4. Delete testpoints.txt file created in step 1

5. Create a database backup using "dfm backup create missing_data"

6. Send us the missing_data.ndb file created in data directory (/opt/NTAPdfm/data on linux)

7. Send us the contents of log directory (/opt/NTAPdfm/log on linux)