2009-11-17 06:45 AM
I am testing SnapMirror Lag Alerting ,
12382 2566 snapmirror:nearly-out-of-date Warning 10 Nov 16:58
12401 2566 snapmirror:out-of-date Error 10 Nov 17:58
I am using Snapmirror Replication Policy in Disaster Recovery ,
I was able to get these events working in protection manager ,with Mirror Policy Error/Warning Thresholds,
But see these working inconsistent ,with Replication Policy,
Do the alerts need to be more than the Replication time of the relationship ?
If these events trigger more than once or only Once and need to be ackowledge before another event for each relationship
I am testing on snapmirror relationship thats being replicated every hour
2009-11-17 09:43 AM
Hi Modugu --
I think you've got two questions.
First, do the lag warning and error thresholds need to be longer than the replication frequency of your relationship? In other words, if you are mirroring once a day, can you set the lag warning threshold at one hour? No, the lag measures how much time has elapsed since the last replication started, so just before the next scheduled update, the lag will be about 24 hours. You'd want to set the lag threshold at something like 25 to 30 hours, depending on how long an update takes.
Second, do you need to acknowledge the events. No, these events automatically reset themselves when the lag changes state. So, if you have a lag warning and the relationship gets updated (so the lag now becomes much shorter), a "Lag OK" event will trigger and that automatically replaces the lag warning event as the "current" event for that relationship. Many DFM events work like this.
2009-11-17 03:14 PM
I am Replicating every Hour and using warning/error thresholds less than the replication time ,
sm.lag events Warning and Errors should trigger every hour as per the Policy.
Although this is not practical in production,I am testing as a scenario to check the event generation.
I dont see sm.lag events everyhour like they should ,
Fewtimes I see the events dont show up in NMC ,and I can only see them with cli, dfm event list ,
I was able to see these alerts in the past ,
I am testing to make sure sm.lag events trigger for all scenarios as per policy definition.
I used the same in Protection Manager , with threshold definitions in Mirror Policy and it worked ,got snapmirror events every hour