Closed Bug 158293 Opened 24 years ago Closed 24 years ago

bad request error (iPlanet Web Server needed upgrade)

Categories

(Core :: Networking: HTTP, defect)

Sun
Solaris
defect
Not set
normal

Tracking

()

VERIFIED INVALID

People

(Reporter: peter.jamieson, Assigned: darin.moz)

Details

Attachments

(3 files)

From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:1.0.0) Gecko/20020530 BuildID: 2002020516 I am requesting a frameset and frames from an iPlanet 4 web server and a bad request occurs. If i truss the browser, the fault goes away which leads me to believe that it's a timing issue (doesn't happen in netscape 6) I managed to get a truss trace by trussing the browser and the web server. Reproducible: Always Steps to Reproduce: 1. Is caused by calling a JSP which is not public but i will attach a truss of the fault for investigation. Actual Results: I got a bad request in one of the frames Expected Results: It seems that the fault is partially with the web server for sending the "bad request" message but it has no header information so mozilla should not display it.
This is a truss of the browser taken when the fault occurs
Attached file Truss of the webserver —
This is a truss taken of the web server when the fault occurs
truss output is basically useless for debugging this sort of problem. IF you suspect an HTTP problem, then setting the env variables: NSPR_LOG_MODULES=nsHttp:5 NSPR_LOG_FILE=bug158293.log and then running mozilla on those pages will produce a log file. IF you attach that, then it could be used ot help track this down. HAve you enabled pipelining, perhaps?
Assignee: new-network-bugs → darin
Component: Networking → Networking: HTTP
QA Contact: benc → tever
Do i need a debug build for this to work. I tried it and couldn't find any log file.
Attached file mozilla log file —
The logging from mozilla of the fault occurring
line 516 of the truss (timestamp 2.3227) shows that the post is being sent in 2 chunks. This is being read by the webserver as 2 requests and causing the fault. Why is the request being sent in 2 chunks?
then the webserver is probably at fault. TCP/IP allows data packets to be divided into smaller packets, so there is never any guarantee about the number of packets that will be sent.
any chance you could provide a packet trace corresponding to the "bad request" ?
The web server is definately at fault. Apparently an HTTP 1.1 thing that requests can be chunked (according to iPlanet webserver release notes) and this was a new feature added to the web server. Installed a service pack for the server and the problem miraculously dissappeared.
Status: UNCONFIRMED → RESOLVED
Closed: 24 years ago
Resolution: --- → INVALID
Verified invalid.
Status: RESOLVED → VERIFIED
QA Contact: tever → junruh
Summary: bad request caused by timing issue. → bad request error (iPlanet Web Server needed upgrade)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: