Loading ...
Sorry, an error occurred while loading the content.

Re: [pcgen_webteam] Hudson Server and other software reboots on the server

Expand Messages
  • Martijn Verburg
    Hi all, I ve started an ADMIN project and have created the first issue there. This project should be used to record issues that need the attention of the
    Message 1 of 1 , Jun 19, 2010
    • 0 Attachment
      Hi all,

      I've started an ADMIN project and have created the first issue there. This
      project should be used to record issues that need the attention of the Admin
      team (e.g. CI server, Website, SF, Y! Groups, Jira etc).

      K

      On Wed, Jun 9, 2010 at 3:14 PM, Martijn Verburg <martijnverburg@...>wrote:

      > Sounds reasonable to me, shall we canvas for any extra thoughts on this on
      > the main list?
      >
      > K
      >
      >
      > On Wed, Jun 9, 2010 at 3:07 PM, Andrew Maitland <drew0500@...>wrote:
      >
      >>
      >>
      >> If we're going to make a new Project might as well make it broad and cover
      >> the Entire Admin team. Since I can't imagine the project will see a lot of
      >> use once we get things stabilized.
      >>
      >> --
      >> Andrew
      >>
      >> ------------------------------
      >> *From:* Martijn Verburg <martijnverburg@...>
      >> *To:* pcgen_webteam@yahoogroups.com
      >> *Sent:* Wed, June 9, 2010 5:18:18 AM
      >> *Subject:* Re: [pcgen_webteam] Hudson Server and other software reboots
      >> on the server
      >>
      >>
      >>
      >> Hi James,
      >>
      >> All sounds reasonable. Makes me think though, do we want to have a JIRA
      >> project for the web team? That way I can have peace of mind knowing that
      >> some of these issues are recorded :).
      >>
      >> K
      >>
      >> On Wed, Jun 9, 2010 at 1:09 PM, James Dempsey <jdempsey@...>wrote:
      >>
      >>>
      >>>
      >>> Hi Kar,
      >>>
      >>> Sadly yes this is a quick fix to a problem which we haven't diagnosed.
      >>>
      >>> I put in the bounce of Hudson to stop the dev list from getting spam
      >>> without spending hours on diagnosing the issue. It could be that it is
      >>> simply that our reports are big, and to help that I have also dropped
      >>> the checkstyle error list by 95% (it wasn't properly configured). It
      >>> could also be a memory leak in Hudson. The other thing we should do is
      >>> to update Hudson to the latest patch, but that will have to wait until I
      >>> have some time.
      >>>
      >>> Regards outage pages, Hudson has a 'starting' screen which will show
      >>> pretty quickly after the bounce, so I expect we are only not covered for
      >>> a minute or two. Given the audience of the CI server (devs) I don't see
      >>> it as critical that there be a neat display there 24/7.
      >>>
      >>> Cheers,
      >>> James.
      >>>
      >>> On 9/06/2010 9:13 PM Martijn Verburg wrote
      >>>
      >>> >
      >>> >
      >>> > Hi Anestis,
      >>> >
      >>> > Thanks for the prompt reply!
      >>> >
      >>> > On 2010/06/09 10:38 AM +0000, karianna03 gave their opinion about the
      >>> > topic of [pcgen_webteam] Hudson Server and other software reboots
      >>> > on...
      >>> > >Hi all,
      >>> > >
      >>> > >I saw some work for this discussed, I'm concerned we're not
      >>> > >discovering the underlying issue(s). Can somebody summarise for me
      >>> > >which apps are mem leaking?
      >>> >
      >>> > We don't know if anything is memory leaking.
      >>> >
      >>> > Apache and mysql are well known for using memory and not releasing
      >>> > it. Hence it's good to occasionally restart them.
      >>> >
      >>> > Playing devil's advocate here, we run Apache Httpd and mysql at my
      >>> > current clients site with no leak/reboots required. We are patched to
      >>> > the latest versions mind you, not sure whether the VPS is at that same
      >>> > patch level there.
      >>> >
      >>> > I suspect the same with Tomcat. However, when Huson fails, JIRA is
      >>> > still usually going strong, so that to me sounds like there may
      >>> > possibly be a memory leak with Hudson.
      >>> >
      >>> > That's why I sent Andrew the URL from the Hudson wiki (I think it
      >>> > was) about how to determine if Hudson had any memory leaks.
      >>> >
      >>> > Again playing devils advocate, we also run Tomcat/Hudson with no
      >>> > leaks, however as before we're at the latest patch level.
      >>> >
      >>> > The last time Hudson failed, I restarted Tomcat, Apache, and mysql,
      >>> > and Hudson started working again.
      >>> >
      >>> > Restarting the various services on the server is also a good way of
      >>> > freeing up memory without needing to reboot the VPS.
      >>> >
      >>> > That I completely agree with, I'm just always wary of bouncing
      >>> > 'production servers/services'. That said, I appreciate we only have
      >>> > limited time and resource to investigate these sorts of things and a
      >>> > midnight bounce is acceptable.
      >>> >
      >>> > Do we need to look into a 'site down for maintenance' type message for
      >>> > our various services?
      >>> >
      >>> > K
      >>>
      >>>
      >>
      >>
      >>
      >>
      >
      >


      [Non-text portions of this message have been removed]
    Your message has been successfully submitted and would be delivered to recipients shortly.