Control panel Updater

[NightStorm]

Is fried... just one more thing in the list, I suppose.
Run the webupdater for a simple thing... to update the API.
It fades the page with the "Loading, please wait..." window thing... server load climbs... climbs a little more... gets a bit higher... then the page returns a DNS error, and fails to load.

top shows that rpmq is chewing up the server pretty good while the page is supposedly updating (emphasis on 'supposedly', of course)... but the update does nothing. API still is not updated.

[root@server psa]# ps -x |grep rpmq
18656 ?        RN     1:47 /usr/lib/rpm/rpmq -q --all --queryformat N:%{NAME}\nF:system_installed_%{NAME}-%{VERSION}-%{RELEASE}.%{ARCH}\nf:%{INSTALLTIME}\nV:%{VERSION}\nL:%{RELEASE}\nA:%{ARCH}\nB:%{BUILDTIME}\nE:%{EPOCH}\n[R:%{REQUIRENAME} %{REQUIREFLAGS:depflags} %{REQUIREVERSION}\n][P:%{PROVIDENAME} %{PROVIDEFLAGS:de
22955 pts/2    SN     0:00 grep rpmq

Now that the page is not loaded, the top shows:

  PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME CPU COMMAND
  490 qmailq    39  19 32676  27M  2400 R N  38.7  2.7   5:40   0 spamd
18656 root      39  19 10884  10M  2768 R N  36.8  1.0   2:47   0 rpmq

Now... while I prefer to use SSH to update, just so I know when exactly Plesk breaks my server, I don't see why they would include a web-based updater that won't even update, but instead leaves the process hanging and times out the webpage... and fails to update.

 

[oucil]

Updater dying..

I was having the same issue and found in another thread that a bunch of guys had drilled the problem down to spamassassin. I stopped the spamassassin service and have since tried the updater again and it seems to be going through.

Worth a shot anyway.

 

[NightStorm]

Yeah, I was in on the spamassassin topic... this problem is not related. Spamassassin and imap are shut down before I run the updater, just to prevent mail messages from being improperly delivered during update, so it's not what is breaking the updater. It's looking like what is breaking it is Plesk timing out while trying to decide what exactly it's going to update... a problem that was faced in the early 7.x days as well, when the server load would go through the roof and nothing would actually update, whenever someone tried to do even the most simple of tasks with the built-in updater in the control panel.

 

[oucil]

I see what you mean, I was lucky however in that I was finally able to get through the update on the upteenth try. I too like to keep an eye on things in top and am getting the same issue with rpmq rumping up into the 90%+ area and hovering around up there.

I should add though now that I notice, that the API was not updated with the rest of it and it was selected. My update did end with a crash but it seams that the API must have been last on the list for everything else worked.

 

[troymc]

I have had this problem ever since upgrading my test box to 8.0. I am on fc2. There were more updates today that I though might fix the problem, but they didn't. I still can update through the shell, just not through the control panel. Any ideas would be appreciated.