i've tried to get this sorted out in the past, with no luck. i'm trying again. here's my problem:
i have a bunch of RoR application servers. each server should have fifteen mongrel_rails processes running at all times. said mongrel_rails processes give details of what they're doing in ps. e.g.: 5 S rails 1081 1 TS 24 - 51373 - 11:33 ? 00:00:11 mongrel_rails [8000/0/31]: idle 5 S rails 1084 1 TS 24 - 52335 - 11:33 ? 00:00:14 mongrel_rails [8001/0/31]: idle 5 S rails 1087 1 TS 24 - 52770 - 11:33 ? 00:00:13 mongrel_rails [8002/0/31]: idle 5 R rails 1090 1 TS 24 - 55668 - 11:33 ? 00:00:18 mongrel_rails [8003/1/30]: handling 10.30.40.121: GET /request/784809532801/0/7619/12330/39 5 S rails 1093 1 TS 24 - 57512 - 11:33 ? 00:00:18 mongrel_rails [8004/0/31]: idle 5 S rails 1096 1 TS 24 - 54646 - 11:33 ? 00:00:16 mongrel_rails [8005/0/31]: idle 5 S rails 1099 1 TS 24 - 54066 - 11:33 ? 00:00:15 mongrel_rails [8006/0/32]: idle 5 S rails 1102 1 TS 24 - 52892 - 11:33 ? 00:00:14 mongrel_rails [8007/0/31]: idle 5 S rails 1105 1 TS 24 - 53299 - 11:33 ? 00:00:13 mongrel_rails [8008/0/31]: idle 5 S rails 1108 1 TS 24 - 55336 - 11:33 ? 00:00:16 mongrel_rails [8009/0/32]: idle 5 S rails 1111 1 TS 24 - 53250 - 11:33 ? 00:00:14 mongrel_rails [8010/0/31]: idle 5 S rails 1114 1 TS 24 - 53704 - 11:33 ? 00:00:18 mongrel_rails [8011/1/30]: handling 10.30.40.121: GET /javascripts/cache/bundle.js 5 S rails 1117 1 TS 24 - 52768 - 11:33 ? 00:00:16 mongrel_rails [8012/0/31]: idle 5 S rails 1120 1 TS 24 - 54946 - 11:33 ? 00:00:16 mongrel_rails [8013/0/30]: idle 5 S rails 1123 1 TS 24 - 52747 - 11:33 ? 00:00:14 mongrel_rails [8014/0/30]: idle i've tried several dozen incantations of the process monitoring dialogue to try to get this to work properly, to no avail. i want zenoss to alert me if one or more of those mongrel_rails processes is *not running*. the problem is, no matter what i do - zenoss interprets activity in the ps output as the process no longer running - for example, the [8011/1/30] instance above - it's a brief blip of activity, but zenoss considers the process in error since it's not listing "[8011/1/30]: idle". i have the process regex as mongrel_rails "ignore parameters" true. zalertonrestart false. zcountprocs true. zfailseverity error. zmonitor true. i don't know python, i've tried understanding the regex rules, but nothing i've tried works. ^mongrel_rails ^mongrel_rails.* mongrel_rails.* i settled on simply mongrel_rails as that's the only part of the string i want interpreted. but it doesn't work. zenoss insists on interpreting the parameters in ps output, and sees a difference between mongrel_rails [8011/1/30]: handling 10.255.10.21: GET /javascripts/cache/bundle.js and mongrel_rails [8011/1/30]: idle and starts alerting like mad, even though i've set it to 'ignore parameters'. i even tried with ignore parameters set to false just to see if it made a difference - it didn't. wtf am i doing wrong here? i have GOT to get process monitoring here to work. i've posted about this before, but either i'm too stupid to understand the explanations, or, well, okay - i admit it: i'm simply too stupid to understand the explanations. help! -------------------- m2f -------------------- Read this topic online here: http://forums.zenoss.com/viewtopic.php?p=31370#31370 -------------------- m2f -------------------- _______________________________________________ zenoss-users mailing list [email protected] http://lists.zenoss.org/mailman/listinfo/zenoss-users
