We just had to reboot each node in a HA cluster one at a time - it takes about an hour.

A once-off maintenance window would have been great - particularly for those who use the recurring windows - a once-off window does not require you to temporarily reconfigure the recurring windows.

I also believe this is a valuable suggestion particularly with the Maintenance windows functionality changes in 5.2. Previously for scheduled maintenance (most wide use case is MS Patching) we would use maintenance windows to eliminate paging for server restarts etc.

Now that maintenance windows only suppress performance related events, we are waking up our on call staff during this maintenance for no reason as this is scheduled maintenance. The only workaround has been suspending monitoring during this time period which results in losing data.

Additionally suspending monitor is not something that can be scheduled as a maintenance window can. If redgate's solution to this is to use the suspend monitoring function can that be added as a schedule?

Agreed. It's more of an alarm suppression rather than a maintenance window but needed regardless.

We have a big application that creates a lot of Blocked Processes during its deployment action.
I would like a way to Trigger a Suspension of Monitoring or Maint Window. i.e API?? Or TSQL INSERT / UPDATE

Thanks for your feedback Joe.

Definitely in agreement about this one. I can point you to another tool that we also use for monitoring various other servers and services that already has this option - Logic Monitor. It has the ability to browse to the correct node (host or instance) and then specify a "Quick Schedule Downtime" event that suspends monitoring for an hour, after which normal monitoring resumes. This is ideal in the situation where you are working on a fault in an instance/host and want to be able to bounce it several times without being hit with the various alerts for failover events, services down etc etc

Thanks for your suggestion Jonathan.

Regards,
Priya

Thanks for the suggestion Bob.

The current workaround for this would be to navigate to Configuration > Alert settings, select the monitored entity using the tree on the left hand side, then click the Long-running query alert and select the radio button 'disabled'. To recieve alerts again, select 'enabled'.

Ideally, how would you want to 'silence' an alert and what kind of configuration would be most useful to you?

Thanks again for the suggestion,
Adam

Allow an alert to be "silenced" temporarily so it won't be raised for a configurable amount of time. Example: if a user is running some particularly long queries, silence them so queries from that user don't trigger that alert for a few hours.