A couple of times now, (once last week, once today) I've come in to find that the web proxy is offline - so the proxy refuses requests for all filtered people.
On the display page the only service with a cross is web filter.
Rebooting the box fixes.
I can't see anything in the logs.
Is there another place where I can restart the web filter service on it's own or find out why this service is falling over?


First thing to check is probably disk usage - that would chime with "nothing in the logs" (nowhere to put it!).
If that's not the case, you could be seeing a case of filterfailitis which has occasionally manifested itself for some unlucky users. Our developers believe that they've isolated the cause, and there should be a fix for it this week in Guardian update (lucky) 13. Would prefer to make sure it isnt something more mundane first. Talk to one of our support lads and find out - but certainly check disk usage first.
Disk usage is not an issue.
I have 58 gigs remaining on the logs partition, about 5% used.
Must be 'filterfailitis'![]()
I think there are a few areas. I've had the clock drift too far out on a VM as ntp stopped working. I think it also dies if the machine account expires, the authentication services stops working and produces lots of errors. Rebooting the DCs here seems to break the connection too (was prompting for passwords this morning). Currently testing the patch that fixes the SEGV issues with FP2.

I've had the segv patch for guardian installed here for a couple of weeks now and haven't seen a recurrence of the web filter dieing so looks like the devs have got it sorted.
Ben


There is a fix for "linux clock drift with VMWare and SMP" - but unless the clock drift is causing problems between NTP updates, I shouldn't worry.
If anyone is interested I will dig out the relevant docco.
There are currently 1 users browsing this thread. (0 members and 1 guests)