Monitor snapmirror lag by DFM

[MARIKOTAGAWA]

Hi, I'm looking for a way to monitor and create an alert when a snapmirror lag exceeds a certain threshold.

Can I setup monitoring snapmirror lag by DFM, that snapmirror relationship has set on the filer and snapmirror has been running?

Can I monitor and create an alert for snapmirror relation whose source filer and destination filer are registered on sperate DFM server?

I setup DFM as follows and set snapmirror threshold for testing. But no snapmirror lag event arose and no alarm created.
Could you review my configuration below and let me know what's wrong, if any?

- Environment:

Snapmirror source:
- Location - Production DC - DC_A
- DFM server - DFM_A (10.40.1.1)
- Filer - Filer_A (10.40.1.2)
- Volume - src_vol1
           src_vol2

Snapmirror Destination:
- Location - DR DC - DC_B
- DFM server - DFM_B (10.50.1.1)
- Filer - Filer_B (10.50.1.2)
- Volume - dest_vol1
           dest_vol2
           
snapmirror relation: (Between Data centers via WAN)

Filer_A:src_vol1 -> Filer_B:dest_vol1  (semi-sync) -- 5min lag
Filer_A:src_vol2 -> Filer_B:dest_vol2 (semi-sync)  -- 23days lag

Since the dest_vol2 volume size is smaller than src_vol2, the snapmirror lag occurs. I wasn't aware of it.

- My Configuration
on DFM_A and DFM_B

1. Setup an alarm for snapmirror lag exceeds 2days
1-1. Management Console - Setup - Alarms - Add
    - Group Section: Global
    - By Event Name: SnapMirror Out of Date
    - SNMP Trap Host: xx.xx.xx.xx
    - Email: mariko@xxx.com

2. Setup an alarm for snapmirror lag exceeds 1.5 days
2-1. Setup - Alarms - Add
    - Group Section: Global
    - By Event Name: SnapMirror Nearly Out of Date
    - SNMP Trap Host: xx.xx.xx.xx
    - Email:mariko@xxx.com

3. Check the current thresholds
    $ sudo dfm option list snapmirrorLagErrorThreshold
    $ sudo dfm option list snapmirrorLagWarningThreshold

4. Modify event severity
   Change the snapmirror:out-of-date severity to Critical snapmirror:nearly-out-of-date severity to Error
    $ sudo dfm eventtype modify -v critical snapmirror:out-of-date
    $ sudo dfm eventtype modify -v error snapmirror:nearly-out-of-date

5. For Test purpose, modify the threshold to 2 hours and 4 hours.

    $sudo dfm option set  snapmirrorLagErrorThreshold=2:00
    $sudo dfm option set  snapmirrorLagWarningThreshold=4:00

I expected alarms will be fired for Filer_A:src_vol2 -> Filer_B:dest_vol2 (semi-sync)  -- 23days lag.
But nothing happened. 

Do you know what's wrong? I appreciate your comments in advance.

Thank you

Mariko
 

[fschuber]

Hi, not sure if it helps but consider the following:

- All Filers that you monitor have to be in the same DFM

- Consider the monitoring intervals, they might be up to 8 hours

[MARIKOTAGAWA]

Thank you fschuber for your comment.

> - All Filers that you monitor have to be in the same DFM

Yes, thi is the issue. 2 filers (source and destination) are not in the same DFM. The filers are in production DC and DR DC, and I installed DFMs in the each DC to monitor filers in the each DC.

I understood that to generate a snapmirror lag alert, filers where source and destination volumes reside need to be in the same DFM.  So in my environment, I can't generate a snapmirror lag alert.