Force re-check of index fragmentation alert
The index fragmentation alert is checked only once per week, because of the load is causes on the system. I am fine with that.
However, I would like to have the option to force an immediate re-check for a specific database - for instance by adding a "re-check" button to the alert window.
SCENARIO: The index fragmentation check runs on Saturday and triggers an alert on one database. On Monday I check the alert but I know that the weekly index defragmentation job runs on Sunday, so it is likely that the problem is already fixed. But I like to be sure (because if not, I really need to investigate what went wrong in the defrag job). Plus, I like to get the alert removed from my dashboard, but if it's not fixed then as far as I know it will not be re-raised when the check sees the same issue next week.
Hugo Kornelis
shared this idea
-
Alex
commented
I also want to see this.. Dont get me wrong "Adam (User Experience Specialist, Red Gate Software)", it will be nice to adjust the time its ran, but that doesnt really do what I need. I agree with OP that there should be a button to recheck the single fragmented indexes. I get these alerts every single week (one reason because of timing) and the only way I can get rid of them is clear the alerts, which is great, but a week later, I get the alert closed email and get confused as to what is actually fragmented and what is not.
I have DBs with 600k+ tables and 3+ indexes per table, so I cannot reforce an entire scan and I shouldnt have to be editing the config file every week just to clear an alert that is no longer a thing.
-
Hi Hugo
If you open the config file in %programdata%\red gate\sql monitor\RedGate.SqlMonitor.Engine.Alerting.Base.Service.exe.settings.config near the bottom you will see this:
<schedule qualifiedChannelName="[Cluster].[SqlServer].[Database].[Table].[Index]">
<weeklySchedule dayOfWeek="Sunday" hour="1" minute="00" />
</schedule>The second line determines when the index fragmentation check runs and can be changed, the comments above it in the file explain the options.
Once that has been edited, restart the Base Monitor.
I hope that helps.
-
Hugo Kornelis
commented
Also, can I edit this idea to extend to a re-check of all alerts?
There are lots of alerts that do not need to be checked very often. For example, I have custom alerts for Agent jobs without notifications set, or for too many or too few VLFs; these all are checked once per day, but I'd like to be able to recheck after fixing what I think is the offender.
-
Hugo Kornelis
commented
@Matthew: Sorry, did not see your response sooner.
How exactly do I configure the date/time for the index fragmentation check? -
Matthew Flatt
commented
Hi Hugo,
Not a solution but it is possible to configure when the index fragmentation check occurs, so it could be moved to after when your weekly maintenance if that might make the alert more useful for the time being?
Matt
