Srx5800 resource leak?

  • Am working with JTAC, and don’t want to go into great detail here, but would just like to ask if this situation sounds at all familiar to anyone.  We have a very lightly used SRX5800, moving from testing toward more production use.  It’s been working fine with perhaps <50 users or sessions for several months.  We recently added another small network, with about 100-200 users.  Everything works fine for awhile, until at some point, within hours or days, apparently dependent on usage since time to failure is as little as an hour during the busiest part of the workday, the system becomes less responsive, and less able to forward traffic, until we have to reboot it to restore service.  Load averages remain low, there are no obvious errors or problems being logged, JTAC hasn’t yet identified the problem.  I know this leaves a hundred questions unanswered, I’m just wanting to know if the general description is recognizable to anyone.  Thanks.

  • 10.3R2.11

  • I have no idea, but if you setup netflow and download a trial netflow colelctor, you should be able to profile the applications/ip and maybe users who are generating traffic.  Their may be a pattern.

    I am using Netflow on my SRX100.

    What Junos software version are you running?