[OSM-dev] api/www flaky tonight, disk problem?

Tom Hughes tom at compton.nu
Mon Jun 25 00:09:35 BST 2007


In message <1182723539.5277.6.camel at localhost.localdomain>
          Jon Burgess <jburgess777 at googlemail.com> wrote:

> The machine running the WWW / API seems to be misbehaving tonight. It
> has been slow for the past few hours and the SMART stats of HDA are not
> looking good. The munin page[1] has the following:
> 
> 
> "This graph shows the value of all S.M.A.R.T attributes of drive hda
> (IBM-DPTA-372050). smartctl_exit_status is the return value of smartctl.
> A non-zero return value indicates an error, a potential error, or a
> fault on the drive.
> ...
> [table]
> smartctl_exit_status smartctl_exit_status gauge 1 None"
> 
> 
> Looks like a non-zero exist status to me, maybe hda is on its way out?

There's no real problem with the web server, other than only having
limited memory which limits the number of fastcgi daemons we can run
to service API requests.

It was just being swamped with requests this evening - there was a
bit of a queue but it seems to have been servicing them as best it
can.

There are no problems reported by smartctl so I don't know what
munin is talking about.

Tom

-- 
Tom Hughes (tom at compton.nu)
http://www.compton.nu/




More information about the dev mailing list