#ONTAP: SnapMirror Dashboard

1 messages · Page 1 of 1 (latest)

winged umbra
#

Do step-by-step instructions currently exist on how to properly configure the ONTAP: SnapMirror Dashboard?

“To monitor both sides of a snapmirror relationship, both source and destination clusters must be monitored by Harvest. Both clusters need to export their metrics to the same Prometheus instance.”

Currently we've got 2 separate NAbox servers (one for the SRC, the other for the DST) collecting metrics against each of our SnapMirror SRC + DST clusters. Is there a way to export just the needed SnapMirror metrics since they're both geographically separated and to minimize WAN network bandwidth impact?

naive leaf
#

@winged umbra Could you share Harvest version?

#

Starting with Harvest version 24.11, the SnapMirror dashboard has been simplified and divided into two separate dashboards as attached. Could you please upgrade to the latest Harvest version (25.02) and check if this resolves your queries?

image.png
winged umbra
naive leaf
#

Please upgrade to 25.02 which is latest version.

winged umbra
naive leaf
#

Nice!

sand stirrup
#

I read the 25.02 release note on github, which said that v25.02 removed the WorkloadDetail and WorkloadDetailVolume templates. Did lun_other_ops metric not show up because it was collected from either of these removed templates?

naive leaf
#

Is only lun_other_ops missing or there are other metrics also?

naive leaf
sand stirrup
#

lun_other_ops, lun_total_ops, lun_total_latency, lun_other_latency, lun_total_data, lun_other_data, and volume_other_data

sand stirrup
naive leaf
sand stirrup
#

In 'conf/keyperf/9.15.0/lun.yaml', does '9.15.0' refer to the Netapp Ontap version?

naive leaf
#

I see, Yes that is the reason. These metrics are available via KeyPerf collector only.

#

You need add this collector in yout Harvest config

#

Could you share which collectors have you configured for the relevant cluster?

sand stirrup
#

Do you mean keyPerf? I already have it. In my 'collectors' section, I have Zapi, ZapiPerf, Ems, REST, and KeyPerf.

naive leaf
naive leaf
#

ZapiPerf lun template doesn't have these metrics listed. Let me check if they are present in ONTAP

#

They are present via ZapiPerf collector. We'll need to modify template. I'll share template with you shortly.

#

Could you try below lun.yaml template instead of https://github.com/NetApp/harvest/blob/main/conf/zapiperf/cdot/9.8.0/lun.yaml for ZapiPerf


name:                     Lun
query:                    lun
object:                   lun

counters:
  - avg_latency                   => total_latency
  - avg_read_latency
  - avg_write_latency
  - avg_xcopy_latency
  - caw_reqs
  - enospc
  - instance_name
  - avg_other_latency             => other_latency
  - other_ops
  - queue_full
  - read_align_histo
  - read_data
  - read_ops
  - read_partial_blocks
  - remote_bytes
  - remote_ops
  - total_data
  - total_ops
  - unmap_reqs
  - vserver_name                  => svm
  - write_align_histo
  - write_data
  - write_ops
  - write_partial_blocks
  - writesame_reqs
  - writesame_unmap_reqs
  - xcopy_reqs

override:
  - writesame_reqs: rate
  - writesame_unmap_reqs: rate
  - caw_reqs: rate
  - unmap_reqs: rate
  - xcopy_reqs: rate


plugins:
  LabelAgent:
    # There are two flavors of lun names
    # /vol/vol_georg_fcp401/lun401
    # /vol/vol_georg_fcp401/lun401/lun401
    split_regex:
      - lun `^/[^/]+/([^/]+)(?:/.*?|)/([^/]+)$` volume,lun

export_options:
  instance_keys:
    - lun # edited by plugin
    - svm
    - volume # added by plugin
#

lun_other_data is not present in ZapiPerf output from ONTAP.

sand stirrup
naive leaf
#

That is right but it only select those objects from KeyPerf which are not present in ZapiPerf.

sand stirrup
#

Okay, to sum up, those metrics I listed above are indeed collected from ZapiPerf, but since they were left out of the lun.yaml file, they were not collected. Is that correct?

naive leaf
#

That is correct.

#

KeyPerf is mostly for ONTAP ASAr2 as ZapiPerf and RestPerf don't work with that.

sand stirrup
#

Thanks a lot for your help. I have one last question. In my harvest.yml file, I only included KeyPerf in the collectors section. I do not see lun_read_ops listed in the conf/keyperf/9.15.0/lun.yaml filr, but on my prometheus, lun_read_ops metric is collected. How does this happen?

naive leaf