Cliff, If you look in the output of "zenactions run -v10", do you have a pipe character proceeding the strings like /Status/Ping in the SQL queries it tries to run to match your alert filters, like mentioned in the bug you referenced? If so, you're running into this: http://dev.zenoss.org/trac/ticket/1085 (which looks exactly like 1048...hmm...)
Example, note the WHERE clause with EventClass : SELECT device,component,message,firstTime,summary,severity, evid FROM status WHERE (eventClass like '%|/Status/Ping%') AND evid NOT IN (SELECT evid FROM alert_state WHERE userid='daviswr' AND rule='Down10') It's affecting me as well (on 1.1.1 and 1.1.2), so at the moment I'm mainly looking at severity in my alert rules, since that's an integer comparison. I've tried digging around in some of the ZenEvents code (zenactions.py, WhereClause.py, maybe ActionRule.py... or I could be completely wrong), but Python isn't one of my strong points. Oddly enough, this wasn't broken in 1.0.2 (I can't speak for 1.1.0). I've seen at least two other threads on here dealing with one, all dead-ending with a link to 1085. Is this happening to anyone else? Wes [EMAIL PROTECTED] wrote:
Hi: I have a base alert (email) rule, and it works fine. I receive email as error events are added and cleared. I added a second email rule, this one with a different (valid) email address. It's configured with the default filters, except it has a "Device Class" filter set to "is /Server/Windows". I am *NOT* receiving email to this address when /Server/Windows error events occur (though I do receive email to the first address). I added a third email rule, with a third (valid) email address, this one configured with the default filters except it has an "Event Class" filter set to "is /Status/Ping". I am not receiving email to this address when /Status/Ping error events occur (though I am notified of them at the first address). As a test, I removed the "Device Class" filter from the second email rule and the "Event Class" filter from the the third. I then rebooted one of my Windows machines and was notified of the failure at all three addresses. I then added the filters back, re-rebooted, and only received one alert. This seems similar to the bug reported in http://dev.zenoss.org/trac/ticket/1048 (which deals with event commands, not alerts). I tried the patch (http://dev.zenoss.org/trac/changeset/4817), but it had no effect. I am using zenoss 1.1.2. --Cliff Krumvieda
-- Wesley Davis Network Engineering James Madison University Massanutten Hall Room 157, MSC 5735 Harrisonburg, VA 22807 Office: (540) 568-5403 Mobile: (540) 830-8216 [EMAIL PROTECTED] _______________________________________________ zenoss-users mailing list [email protected] http://lists.zenoss.org/mailman/listinfo/zenoss-users
