We have added customized wording to our alert messages for cdm, processes and ntservices. The clear messages do not contain the customized wording. Any ideas on how to keep the customized wording in the associated clear messages?
That varies from probe to probe...
In the cdm probe, I do not see a way to customize clear messages at all. That does not appear to be an option in the GUI, and no clear messages exist in the config file where they could be easily tweaked. Did you find a way to do this?
In the processes probe, I see ways to customize the default clear message and create new clear messages. However, I do not see a way to configure a profile to use a non-default clear message even though other non-default messages can be selected in the profile. It looks like you can override the clear token, but messages with a level of clear do not appear in the list of options. I suspect you could set the clear token to use a different message in the GUI and then change it to a custom clear message in Raw Configure.
In the ntservices probe, I see ways to customize the default clear message, create new clear messages, and set each profile to a non-default clear message. It should be possible to do all of this in the GUI.
So what exactly did you do to change the clear messages in these probes, and what was the result for each?
I'm looking at this problem now too.
The issue for us is that the clear message doesn't really indicate which original alarm it was from. The only way I can see around this is to make a different watcher profile for every single alarm condition for every process and include the $watcher variable in the clear. This means we wouldn't be able to do things like put process down and process count in the same watcher.
Am I missing something here? Is there a reason why there isn't a clear token for every single alarm message? Or why this doesn't handle it like cdm and send a clear with the original alarm message in it by default?
From what I can tell unless the processes probe is set up to be able to understand a process_down_clear token and relate that to the process_down, then there is no way to do this. Is that right?
Look into preprocessing scripts in the nas. With this you can write LUA in order to manipulate the messages pretty much however you wish.
For those messages where you need to determine additional information than is in the message, many times that information is in the suppression id. For the CDM disk message for instance, the message says that the event is clear but the suppression id contains the drive letter/path that the alarm applies to.
You can use Dr. Nimbus to watch the messaging too
Retrieving data ...