alert notifications
grambowk
Posts: 8
The log shipping monitoring tool is great. However I cannot seem it configure alert notifications.
Is it possible to configure alert notifications to send emails on the occurrence of an alert? If not I would suggest that this is possibly one of the most important features that I would want - and I'm sure others would find this useful.
Without alert notifications I would have to be constantly checking up on the monitoring tool to see if an alert has been raised and naturally this isn't possible nor practical.
Currently, if one of the log shipping jobs fail I can receive an alert. However, there are instances where a job doesn't fail (e.g. the job is running for too long). This can be frustrating because log shipping could end up several hours out of sync before I notice.
Regards,
Karl
Is it possible to configure alert notifications to send emails on the occurrence of an alert? If not I would suggest that this is possibly one of the most important features that I would want - and I'm sure others would find this useful.
Without alert notifications I would have to be constantly checking up on the monitoring tool to see if an alert has been raised and naturally this isn't possible nor practical.
Currently, if one of the log shipping jobs fail I can receive an alert. However, there are instances where a job doesn't fail (e.g. the job is running for too long). This can be frustrating because log shipping could end up several hours out of sync before I notice.
Regards,
Karl
Comments
Information on the current functionality of alerts is covered in the wiki article on the labs:
http://labs.red-gate.com/index.php/Log_ ... and_Alerts
Unfortunately, it is not possible to configure alerts to send email notifications, but it is something we are considering for a subsequent update of the Log Shipping Monitor.
If you have any thoughts on what functionality you would like to see from such a feature, or some indication on what scenarios you would use this for (such as notification of long-running jobs, as you mentioned), we can include your feedback when we look into the next update of the Log Shipping Monitor.
Thanks,
Jason
Functionality that gives the ability to specify an smtp server and the option "send email notification on alert" (true or false), along with an email recipient would be great.
In this instance then, any configured alerts would generate an email that is sent to the supplied recipient if the option is set to true.
In addition to this it would probably be useful to include a "Delay Between Notifications" option where you can provide the number of minutes delay that you want to allow for between the same alert (similar to the Alerts in SQL Server).
The above functionality would be a great start and is all I'd really need but I imagine you could extend it even further.
For example, when specifying alerts you could add a couple of parameters. I'll use the lastrestore alert as an example:
LASTRESTORE {restore type}, {interval}, {send notification}, {notify recipient}
Where {send notification} is a bit value and {notify recipient} is the email recipient.
This would then allow for multiple alerts to be configured, some of which are set to notify someone and some of which might not.
The most likely scenario that I'd use this functionality for is to report on log shipping going out of sync. In our current environment it is important that the latency of log shipping is kept to a minimum at all times. Typically, a log backup or restore will not take any longer than 30 minutes. So in the event that the LASTRESTORE interval is greater than 30 minutes I'd want to investigate the situation immediately. Aside from the obvious need to keep our secondary databases in sync in case we ever needed to failover, when things go wrong with log shipping it is helpful to tackle the problem sooner rather than later.
Hope that helps. If you need any further details please let me know.
Great tool by the way!
Karl
Yes, the standard GUI contains SMTP alerting (I'm using v5.3), however, I am finding that log shipping on large DBs often hangs within the SQL Agent Job itself - this appears to coincide with full SQL backups being performed early in the morning - and the SQL job just continues to report "Executing..." The problem here is that the Redgate job does not report a warning or failure, and hence no alert is sent from the GUI.
I can fix the problem with a restart of the SQL Agent Service (SQL 2000), but I have to manually check the status to discover the problem. An alert from the Log Shipping Monitor if there has been no activity for a certain amount of time would be a more all encompassing way to monitor the process in my opinion.
Thanks!