Exception within ZServer hangs Zope
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, Zope users! Has anyone seen this? I causes zope to just hang, and I don't have a clue as to where to start looking. It seems it happens within the ZServer, but I couldn't find a line in the entire tree of Zope that would assign a suspicious value to that 'p' variable. Can anyone suggest how to detect this and at least restart Zope, as this is causing us service outages (the question is really how to properly programatically restart Zope, as if you push Restart Zope within the control panel). Is killing -15 zope processes safe? 2003-11-25T11:05:58 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 192.168.0.56:1598 at 0x924a2b4 channel#: 124 requests:1> (exceptions.AttributeError:'unicode' object has no attribute 'more' [/var/home/jure/zope/ZServer/medusa/asyncore.py|poll|107] [/var/home/jure/zope/ZServer/medusa/asyncore.py|handle_write_event|402] [/var/home/jure/zope/ZServer/medusa/asynchat.py|handle_write|146] [/var/home/jure/zope/ZServer/medusa/asynchat.py|initiate_send|208] [/var/home/jure/zope/ZServer/medusa/asynchat.py|refill_buffer|195]) Traceback (most recent call last): File "/home/jure/zope/z2.py", line 913, in ? asyncore.loop() File "/var/home/jure/zope/ZServer/medusa/asyncore.py", line 200, in loop poll_fun (timeout, map) File "/var/home/jure/zope/ZServer/medusa/asyncore.py", line 111, in poll obj.handle_error() File "/var/home/jure/zope/ZServer/medusa/http_server.py", line 440, in handle_error asynchat.async_chat.handle_error (self) File "/var/home/jure/zope/ZServer/medusa/asyncore.py", line 425, in handle_error self.close() File "/var/home/jure/zope/ZServer/HTTPServer.py", line 325, in close p.more() # free up resources held by producer AttributeError: 'unicode' object has no attribute 'more' - -- Jure Koren, n.i. jure@aufbix.org, GPG pubkey @ http://aufbix.org/~i/public.key -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.0 (GNU/Linux) iD8DBQE/wyLV9iFCvmuhrCIRAn/+AJ42J0XnVo9oeSq62BM+RnzH7EpmtgCgpFZV aXPTBG49Obl7Rj11o2N/KUA= =HwGb -----END PGP SIGNATURE-----
Jure -- What Zope are you using? what OS? which python? what machine? That information is need to help sort out the problem. Is this something that just started happening in a system that has been running for a while without a problem? Have you recently upgraded either Python or Zope? The exception happens in the guts of ZServer during a close when a unicode string is found on a fifo. I suspect that changing the test to allow ignore unicode stings as well as traditional strings would solve the problem, but I don't know the code well enough to say for certain. On Tue, 25 Nov 2003, Jure Koren wrote:
Has anyone seen this? I causes zope to just hang, and I don't have a clue as to where to start looking. It seems it happens within the ZServer, but I couldn't find a line in the entire tree of Zope that would assign a suspicious value to that 'p' variable. Can anyone suggest how to detect this and at least restart Zope, as this is causing us service outages (the question is really how to properly programatically restart Zope, as if you push Restart Zope within the control panel). Is killing -15 zope processes safe?
participants (2)
-
Dennis Allison -
Jure Koren