VPS reboot -- changes causing trouble?

thingwarbler

thingwarbler

Perch
Since last weekend's reboot I've been seeing weird IIS crashes caused by ASP application pools failing repeatedly and eventually pulling down the entire system. My server had been running rock solid for weeks until the reboot, so I'm wondering if y'all have any thoughts on what might have changed in connection with the reboot? Any patches etc. that might be causing this grief? The only ASP application I'm running is Smartermail 4.1 Pro.
 
hrmmm, no reports of anyone else, I will go back through and check all the updates run to see if any are IIS related.

Jodohost is on that server now, and I have not seen the pool for it crash.
 
Stephen,

Thx for finding that. I'll get the ticket set up. Just in case others are seeing similar behavior, here's what I'm getting:

3:55:25PM

A process serving application pool 'plesk(default)(pool)' suffered a fatal communication error with the World Wide Web Publishing Service. [...]
nk/events.asp.

Repeat 9 times, then finally a little more than 10 minutes later:

4:07:47PM

Application pool 'plesk(default)(pool)' is being automatically disabled due to a series of failures in the process(es) serving that application pool.

And that's the end of the web server until you go in and manually restart IIS. Then you get a few hours of decent performance and then it starts all over again...
 
Stephen,

Rebooted to let the un-patch take effect, but I don't think it solves anything. I'm still seeing really poor performance now -- every few minutes something causes the application pool error that I described earlier, and eventually IIS comes crashing down.

I haven't changed anything significant lately, so I have to assume that it was something that was tweaked with that last maintenance. While I stay busy explaining to my customers that constant crashes is, in fact, considered a feature of a quality windows server, could you take a second to think about anything else that might have happened?

Couple of leads to go by: this seems to most significantly impact a couple of joomla sites I've got on the server (php, mysql) and not the straight HTML and CF sites.

thx
 
hello,

I have checked and not seeing any other update, has smartermail stopped crashing?

nothing we did changed php or IIS :( I have triple checked that now.

One think yuo may consider is to move php to CGI instead of ISAPI, so even if some code on a side was crahsing PHP, it woudl not crash the pool like ISAPI php can.
 
Stephen,

Thx for the quick reply as always. It wasn't just smartermail that crashed (it turns out), but rather the two application pools handling everything on the server (one is strictly smartermail's web UI, the other (default plesk) is "everything else" -- i.e. all the sites on the box).

The errors I see in the event log are from both pools -- mostly the default plesk pool. The skimpy error logging doesn't give me much to go by in the way of debugging -- AFAIK there's no way for me to see which particular site threw the spanner in the works.

I'll try to switch the php setup, but I seem to recall there was some very compelling reason why I set things up the way they are now (i.e. apps not compatible with php run under CGI because of Windows non-threading issues).
 
You must log in or register to reply here.
Back
Top