ZServer uncaptured python exception
Running logtail on a ZEO client, I am seeing lots of these. And our two public clients are pegging the server at 99% almost continuously. Not sure why. 2008-04-01T09:54:09 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56350 at 0xb63d7a6c channel#: 800 requests:4> (socket.error:(104, 'Connection reset by peer') [/usr/local/lib/python2.3/asynchat.py|initiate_send|218] [/var/www/zope/zope275c2/lib/python/ZServer/medusa/http_server.py|send|417] [/usr/local/lib/python2.3/asyncore.py|send|337]) ------ 2008-04-01T09:54:17 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56590 at 0xb26898ac channel#: 890 requests:> (socket.error:(104, 'Connection reset by peer') [/usr/local/lib/python2.3/asynchat.py|initiate_send|218] [/var/www/zope/zope275c2/lib/python/ZServer/medusa/http_server.py|send|417] [/usr/local/lib/python2.3/asyncore.py|send|337]) ------ Allen
----- Original Message ----- From: "Allen Schmidt Sr." <aschmidt@fredericksburg.com> To: "Zope Users" <zope@zope.org> Sent: Tuesday, April 01, 2008 10:11 AM Subject: [Zope] ZServer uncaptured python exception
Running logtail on a ZEO client, I am seeing lots of these. And our two public clients are pegging the server at 99% almost continuously. Not sure why.
2008-04-01T09:54:09 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56350 at 0xb63d7a6c channel#: 800 requests:4> (socket.error:(104, 'Connection reset by peer') [/usr/local/lib/python2.3/asynchat.py|initiate_send|218] [/var/www/zope/zope275c2/lib/python/ZServer/medusa/http_server.py|send|417] [/usr/local/lib/python2.3/asyncore.py|send|337]) ------ 2008-04-01T09:54:17 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56590 at 0xb26898ac channel#: 890 requests:> (socket.error:(104, 'Connection reset by peer') [/usr/local/lib/python2.3/asynchat.py|initiate_send|218] [/var/www/zope/zope275c2/lib/python/ZServer/medusa/http_server.py|send|417] [/usr/local/lib/python2.3/asyncore.py|send|337])
I am not sure that the error messages you are seeing are related to your high cpu utilization. There is a discussion thread on your error here: http://aspn.activestate.com/ASPN/Mail/Message/zope-list/3492748 If you google 'debug spinning zope' you will find links on how to track down cpu consumption. HTH Jonathan
That makes sense...but I think they are still related, sort of. The thread you offered says the error comes from clients disconnecting...hitting the stop button or loading a different page...effectively doing the same thing. But they are hitting the stop button BECAUSE the site is slow and spinning. I have Deadlockdebugger and that shows a variety of things. Mostly that all 4 threads are always full. I will keep looking. Allen Jonathan wrote:
----- Original Message ----- From: "Allen Schmidt Sr." <aschmidt@fredericksburg.com> To: "Zope Users" <zope@zope.org> Sent: Tuesday, April 01, 2008 10:11 AM Subject: [Zope] ZServer uncaptured python exception
Running logtail on a ZEO client, I am seeing lots of these. And our two public clients are pegging the server at 99% almost continuously. Not sure why.
2008-04-01T09:54:09 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56350 at 0xb63d7a6c channel#: 800 requests:4> (socket.error:(104, 'Connection reset by peer') [/usr/local/lib/python2.3/asynchat.py|initiate_send|218] [/var/www/zope/zope275c2/lib/python/ZServer/medusa/http_server.py|send|417] [/usr/local/lib/python2.3/asyncore.py|send|337]) ------ 2008-04-01T09:54:17 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56590 at 0xb26898ac channel#: 890 requests:> (socket.error:(104, 'Connection reset by peer') [/usr/local/lib/python2.3/asynchat.py|initiate_send|218] [/var/www/zope/zope275c2/lib/python/ZServer/medusa/http_server.py|send|417] [/usr/local/lib/python2.3/asyncore.py|send|337])
I am not sure that the error messages you are seeing are related to your high cpu utilization.
There is a discussion thread on your error here: http://aspn.activestate.com/ASPN/Mail/Message/zope-list/3492748
If you google 'debug spinning zope' you will find links on how to track down cpu consumption.
HTH
Jonathan _______________________________________________ Zope maillist - Zope@zope.org http://mail.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope-dev )
Allen Schmidt Sr. wrote:
Running logtail on a ZEO client, I am seeing lots of these. And our two public clients are pegging the server at 99% almost continuously. Not sure why.
2008-04-01T09:54:09 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56350 at 0xb63d7a6c channel#: 800 requests:4> (socket.error:(104, 'Connection reset by peer')
This exception is annoying in that it's pretty meaningless. Clients disconnecting midway through may be quite a normal thing and it's annoying there's no way to catch this exception if you want to. In your case, I'd guess that your load has recently gone way up (maybe a search engine spider, some other spider of people with M$ Office installed hammering your site for webdav requests) but I wouldn't worry about these... cheers, Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
Thanks Chris. Seeing tons of squid messages about the new msnbot....and it is not following our robots.txt directive. In fact, I have even disallowed it completely and it still goes where it wants. And the updated version was supposed to have been released end of last week....about the time our zeo clients started pegging 99%. Chris Withers wrote:
Allen Schmidt Sr. wrote:
Running logtail on a ZEO client, I am seeing lots of these. And our two public clients are pegging the server at 99% almost continuously. Not sure why.
2008-04-01T09:54:09 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected 8.7.96.136:56350 at 0xb63d7a6c channel#: 800 requests:4> (socket.error:(104, 'Connection reset by peer')
This exception is annoying in that it's pretty meaningless. Clients disconnecting midway through may be quite a normal thing and it's annoying there's no way to catch this exception if you want to.
In your case, I'd guess that your load has recently gone way up (maybe a search engine spider, some other spider of people with M$ Office installed hammering your site for webdav requests) but I wouldn't worry about these...
cheers,
Chris
On Apr 2, 2008, at 11:11 , Allen Schmidt Sr. wrote:
Thanks Chris. Seeing tons of squid messages about the new msnbot....and it is not following our robots.txt directive. In fact, I have even disallowed it completely and it still goes where it wants. And the updated version was supposed to have been released end of last week....about the time our zeo clients started pegging 99%.
Trusting that bots actually care about or respect robots.txt won't get you very far. If all else fails I suggest you look at your logs to find out what IPs or IP ranges those bots come from and block them at the firewall level, using e.g. iptables if you're on Linux. This is a very simplistic invocation: iptables -A INPUT -s <source-IP> -j DROP At least that way you gain some breathing room to come up with a less constrictive solution. jens
Jens Vagelpohl wrote:
Trusting that bots actually care about or respect robots.txt won't get you very far. If all else fails I suggest you look at your logs to find out what IPs or IP ranges those bots come from and block them at the firewall level, using e.g. iptables if you're on Linux. This is a very simplistic invocation:
iptables -A INPUT -s <source-IP> -j DROP
At least that way you gain some breathing room to come up with a less constrictive solution.
Also very effective against errant Google Search Appliances in intranet situtations ;-) Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
I have had similar instances with GSA on my Zope server in that it would try to index something like /mydir/mydir/mydir/mydir/somefile.html Often I've found the case is an href that references ../somethingelse.html I don't know if this might be related to your case but you might want to check. I had to finally end up using the iptables rule already suggested and this was our campus GSA. After talking to the owner of the GSA he told me he found a plugin or fix for Plone for the GSA and we are going to try that to see if it helps with Zope. But this is local so you probably wouldn't be able to ask outside entities to use this. Thomas On Wednesday 02 April 2008 05:50, Chris Withers wrote:
Jens Vagelpohl wrote:
Trusting that bots actually care about or respect robots.txt won't get you very far. If all else fails I suggest you look at your logs to find out what IPs or IP ranges those bots come from and block them at the firewall level, using e.g. iptables if you're on Linux. This is a very simplistic invocation:
iptables -A INPUT -s <source-IP> -j DROP
At least that way you gain some breathing room to come up with a less constrictive solution.
Also very effective against errant Google Search Appliances in intranet situtations ;-)
Chris
-- ==================================================================== Thomas McMillan Grant Bennett Appalachian State University Operations & Systems Analyst P O Box 32026 University Library Boone, North Carolina 28608 (828) 262 6587 "... using OpenOffice.org, and save them back to disk automatically, in MS Word format. They surf the Web, check e-mail, do instant messsaging, view YouTube videos, visit their Facebook pages, learn touch-typing skills and lots more. Our public library has been offering these Linux public stations for the past three years." - Phil Shapiro Linux Journal January 2008 Library Systems Help Desk: https://www.library.appstate.edu/help/ ====================================================================
Thomas Bennett wrote:
I have had similar instances with GSA on my Zope server in that it would try to index something like /mydir/mydir/mydir/mydir/somefile.html
Often I've found the case is an href that references ../somethingelse.html
Yes, that's bad coding on your part. cheers, Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
I absolutely agree and I have preached to the WEB page editors here about this but most of them have no idea of the difference between a relative link, an absolute link, and trash like this. Using Dreamweaver they just know "it works" at least for them. ;-) Thomas On Monday 07 April 2008 06:06, Chris Withers wrote:
Thomas Bennett wrote:
I have had similar instances with GSA on my Zope server in that it would try to index something like /mydir/mydir/mydir/mydir/somefile.html
Often I've found the case is an href that references ../somethingelse.html
Yes, that's bad coding on your part.
cheers,
Chris
-- ==================================================================== Thomas McMillan Grant Bennett Appalachian State University Operations & Systems Analyst P O Box 32026 University Library Boone, North Carolina 28608 (828) 262 6587 "... using OpenOffice.org, and save them back to disk automatically, in MS Word format. They surf the Web, check e-mail, do instant messsaging, view YouTube videos, visit their Facebook pages, learn touch-typing skills and lots more. Our public library has been offering these Linux public stations for the past three years." - Phil Shapiro Linux Journal January 2008 Library Systems Help Desk: https://www.library.appstate.edu/help/ ====================================================================
Allen Schmidt Sr. wrote:
Thanks Chris. Seeing tons of squid messages about the new msnbot....and it is not following our robots.txt directive. In fact, I have even disallowed it completely and it still goes where it wants. And the updated version was supposed to have been released end of last week....about the time our zeo clients started pegging 99%.
Just block the damn thing by useragent in Squid or Apache :-S Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
participants (5)
-
Allen Schmidt Sr. -
Chris Withers -
Jens Vagelpohl -
Jonathan -
Thomas Bennett