Bug 744684 - Traceback shown while registering and server having "Internal Server Error"
Summary: Traceback shown while registering and server having "Internal Server Error"
Keywords:
Status: CLOSED WONTFIX
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: rhn-client-tools
Version: 6.2
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: rc
: ---
Assignee: Milan Zázrivec
QA Contact: Red Hat Satellite QA List
URL:
Whiteboard:
Depends On:
Blocks: 691780
TreeView+ depends on / blocked
 
Reported: 2011-10-10 07:50 UTC by Pavel Holica
Modified: 2011-12-08 19:46 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2011-12-08 19:46:06 UTC
Target Upstream Version:


Attachments (Terms of Use)
abrt.log containing traceback (1.48 KB, text/plain)
2011-10-10 07:50 UTC, Pavel Holica
no flags Details

Comment 1 Milan Zázrivec 2011-10-10 10:30:56 UTC
Internal server error is in truth a problem on the server side.

When the client receives HTTP/500 response, there's not much it can
do about it -- other than display it.

May I suggest to reach out to RHN hosted folks (or admins) to investigate
the issue further (seeing this is an rhn.qa environment, problems like this
are nothing unusual).

Comment 2 Pavel Holica 2011-10-10 10:48:35 UTC
But there is no error message, there is traceback. User shouldn't see any traceback.

Comment 3 Jan Pazdziora 2011-11-24 13:29:31 UTC
If I set serverURL in /etc/sysconfig/rhn/up2date to point to URL which always returns 500 (ISE), I get a nice error message "We could not contact Satellite or Proxy at ...", which comes from _chooseServerPageApply.

So if the server returns 500 (ISE) in consistent manner, the user will see a nice error message right at the beginning.

If the first page passes thou, we don't catch the exceptions much beyond this point and leave it on firstboot to handle it, thus the tracebacks.

I am not sure the effort to address all the places where the exception can appear and to QA it is worth the benefit, with subscription manager being available.

I therefore propose WONTFIX, while acknowledging that we would probably be able to address it properly given we spend enough man hours on the issue.

Mike, could you give the product management's ruling on this bug?

Comment 4 Mike Khusid 2011-11-30 19:57:06 UTC
Agree with proposed resolution in comment 3.

Comment 5 Jan Pazdziora 2011-12-01 10:59:41 UTC
Setting devel nack based on comment 4.


Note You need to log in before you can comment on or make changes to this bug.