Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

[10:50:04 CST(-0600)] * Topic is 'http://uportal.pastebin.com/ - http://www.ja-sig.org/wiki/display/UPC/uportal+IRC+Logs' set by EricDalquist on 11:32:13 CST(-0600)
[10:56:40 CST(-0600)] <EricDalquist> testing
[10:58:24 CST(-0600)] * Topic is 'http://uportal.pastebin.com/ - http://www.ja-sig.org/wiki/display/UPC/uportal+IRC+Logs&#039; set by EricDalquist on 10:58:24 CST(-0600)
[18:13:14 CST(-0600)] <bjagg> Hey guys, tried rolling out 3.2.4. Strange behavior where initial interaction is fine, but if left alone for 30 minutes, all portlets timeout
[18:13:58 CST(-0600)] <bjagg> Layout is fine, but portlets from inframe to proxies display rendering error
[18:14:31 CST(-0600)] <EricDalquist> any other exceptions?
[18:15:10 CST(-0600)] <bjagg> Some jk connection errors
[18:15:54 CST(-0600)] <EricDalquist> nothing weird in the uportal log though?
[18:16:37 CST(-0600)] <bjagg> checking ...
[18:17:22 CST(-0600)] <bjagg> a bunch of java.lang.IllegalStateException: Cannot create a session after the response has been committed
[18:21:09 CST(-0600)] <EricDalquist> huh, thats what you'll see if a portlet is still executing after the portal stopped waiting for it
[18:21:15 CST(-0600)] <EricDalquist> weird that the portal timed out everything though
[18:23:07 CST(-0600)] <bjagg> yep,and only after several minutes. Wonder if cache timeout is involved ... checking
[18:23:35 CST(-0600)] <EricDalquist> the portal should still function correctly even if all caches have all elements removed
[18:23:44 CST(-0600)] <bjagg> ok
[18:26:42 CST(-0600)] <EricDalquist> the other thing you can do is a thread dump
[18:26:54 CST(-0600)] <EricDalquist> and see if the ChannelManager threads are stuck doing something weird
[18:27:07 CST(-0600)] <bjagg> ok cool – thx for the advice!
[18:27:15 CST(-0600)] <EricDalquist> they should just all be waiting if there is no traffic
[18:28:17 CST(-0600)] <bjagg> Are you guys still using Solaris? That's a difference between our test and prod environments
[18:28:46 CST(-0600)] <EricDalquist> yeah
[18:28:49 CST(-0600)] <EricDalquist> we use solaris everywhere
[18:29:03 CST(-0600)] <bjagg> That's reassuring