Closed Bug 570793 Opened 16 years ago Closed 16 years ago

workermgmt app is erroring out

Categories

(mozilla.org Graveyard :: Server Operations, task)

All
Other
task
Not set
major

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: morgamic, Unassigned)

Details

Right now we're seeing: An error has occurred... * Application failed during request deserialization: no element found at line 1, column 0, byte -1 at /usr/lib64/perl5/vendor_perl/5.8.8/x86_64-linux-thread-multi/XML/Parser.pm line 187 The error is likely: 12:05 <skeen> the php code look for ''faultString'' in the response xml and echo's it out This is the old code, so wondering if Dave can help debug this error on the Bugzilla side? Another alternative would be to fast-track the new version of this and get that out there (more in bug 564265).
Severity: minor → major
Any update here?
< morgamic> mrz-work: can someone look at bug https://bugzilla.mozilla.org/show_bug.cgi?id=570793 ? < morgamic> nobody can fill out newhire stuff because the auth portion of the bugzilla API is returning an error < morgamic> Application failed during request deserialization: no element found at line 1, column 0, byte -1 at /usr/lib64/perl5/vendor_perl/5.8.8/x86_64-linux-thread-multi/XML/Parser.pm line 187 < morgamic> code isn't in SVN either < morgamic> afaik < morgamic> so if someone can tar it up and send to me i can look < morgamic> we also had a replacement for the old one; worse case we could swap them out - https://bugzilla.mozilla.org/show_bug.cgi?id=564265 < morgamic> need update on that bug too < morgamic> workermgmt bug re: perl error is https://bugzilla.mozilla.org/show_bug.cgi?id=570793
I've got a new version of this whole thing in bug 564265 ready to go too. I can't debug unless I have the code. Sam said he'd look at it tonight.
code is in svn, I put it there myself before I handed off to skeen a while back. There were two branches, one was the current live site (which was trunk) and one was dtran's aborted rewrite (which was branches/jquery-rewrite). svn+ssh://svn.mozilla.org/moco/webtools/workermgmt/trunk
does this use the bugs-api server or does it use bugzilla's xmlrpc directly?
Dunno, but Sam said: 21:47 <skeen> hey mike , just sent an email with what i found on workermgmt 21:48 <skeen> pretty much looks like something changed serverside behind https://bugzilla.mozilla.org/xmlrpc.cgi
Sam's email had: have there been any recent upgrades to bugzilla? # The login page of the app posts to: https://bugzilla.mozilla.org/xmlrpc.cgi # with the post data (with my pw stared out): <?xml version="1.0" encoding="iso-8859-1"?> <methodCall> <methodName>User.login</methodName> <params> <param> <value> <struct> <member> <name>login</name> <value> <string>skeen.bugzilla@gmail.com</string> </value> </member> <member> <name>password</name> <value> <string>***************</string> </value> </member> </struct> </value> </param> </params> </methodCall> # and is currently getting a resp: HTTP/1.1 200 OK Date: Wed, 09 Jun 2010 04:27:14 GMT Server: Apache SOAPServer: SOAP::Lite/Perl/0.711 Content-Length: 453 X-Backend-Server: pm-app-bugs05 Content-Type: text/xml <?xml version="1.0" encoding="UTF-8"?><methodResponse><fault><value><struct><member><name>faultString</name><value><string>Application failed during request deserialization: no element found at line 1, column 0, byte -1 at /usr/lib64/perl5/vendor_perl/5.8.8/x86_64-linux-thread-multi/XML/Parser.pm line 187 </string></value></member><member><name>faultCode</name><value><string>Client</string></value></member></struct></value></fault></methodResponse>
When exactly did this start? As far as I know, there haven't been any recent bmo upgrades. http://groups.google.com/group/mozilla.support.bugzilla/browse_thread/thread/8e08f20532a8c5c6 Maybe related: https://bugzilla.gnome.org/show_bug.cgi?id=447102
First saw it yesterday.
Second seems x64 related. Did we move bugzilla to a 64-bit box?
skeen: does your bugzilla password contain any weird characters that could be causing issues when being sent via xmlrpc?
Has workermgmt worked since bmo was moved behind zeus a few weeks ago?
Yeah, we've seen a number of new hire requests after..so it has been working. Mike, Dave says we moved Bugzilla to 64 bit boxes a few months back.
Well... it fails now and no code has changed. Do we know what modules have changed over the last week? Server software updates?
Also, if I just use "user" and "pass" as the combination we get the same error. So it's gotta be some bad redirect that causes the loss of the POSTDATA then sends a post with no data in it (or maybe even a bad req). Prod form's action is: <form method="post" action=""> Is that right?
I don't know what else to test except to watch the net traffic on the box to catch what bugzilla is receiving (if anything) and what the intranet box is sending.
This appears to have been broken by an update to the SOAP::Lite perl module. One of the three webservers picked up the new version on June 4th, the other two got it on the 8th. new version: perl-SOAP-Lite-0.711-1.el5.rf reverted to: perl-SOAP-Lite-0.710.08-1.el5.rf Can you try it again?
Works now. Yay!
Status: NEW → RESOLVED
Closed: 16 years ago
Resolution: --- → FIXED
(also, this explains why the errors were intermittent for me on Monday)
We have a winner. It's even a known issue upstream: https://bugzilla.mozilla.org/show_bug.cgi?id=569385
Product: mozilla.org → mozilla.org Graveyard
You need to log in before you can comment on or make changes to this bug.