tracker issue : CF-3304743

select a category, or use search below
(searches all categories and all time range)
Title:

Runtime processes the wrong template for a given request

| View in Tracker

Status/Resolution/Reason: Closed/Withdrawn/Duplicate

Reporter/Name(from Bugbase): Mark Butler / Mark Butler (butlerm)

Created: 08/07/2012

Components: Core Runtime

Versions: 10.0

Failure Type: Non Functioning

Found In Build/Fixed In Build: Final /

Priority/Frequency: Critical / Most users will encounter

Locale/System: English / Linux All

Vote Count: 0

Duplicate ID:	CF-3227975

Problem Description:  Base frameset contains two template URLs, one for each frame. On load, browser submits requests to server.  On execution, the runtime often runs a different template than was specified for one of the requests, generally by substituting the template and parameters of the other simultaneously executing request, as if it had trouble distinguishing between the two.  This never occurred with identical code on CF7, 8, 9.

Steps to Reproduce: Erratic - happens several times a day with (near) simultaneously executing requests from the same client. Requests are HTTPS.  Occurs with both IE9 and Chrome, presumably others.

Actual Result:  Sometimes the same template with the same parameters is executed for each request, even though the URLs (and parameters) submitted were completely different. 

Expected Result:  Each request executes and returns results for the specified template with the specified parameters.

Any Workarounds:  None.

----------------------------- Additional Watson Details -----------------------------

Watson Bug ID:	3304743

External Customer Info:
External Company:  
External Customer Name: butlerm
External Customer Email:  
External Test Config: My Hardware and Environment details:



ColdFusion 10.0 Linux x86 64 bit stock, no updates

Apache web server (version that comes with OS), mod_jk connector

CentOS 6.3 x86 64 bit (Essentially the same as RHEL 6.3 x86 64 bit).  

Intel E3-1240v2 Processor (Quad core Xeon, 3.40GHz 8MB Cache)

Intel server system, 16 GB RAM,  1 TB SATA RAID.

Attachments:

  1. August 09, 2012 00:00:00: 1_cust_inv_payment_2.txt
  2. August 09, 2012 00:00:00: 2_debug_info.txt

Comments:

Can you share a sample code to repro the issue ? Meanwhile I am trying to repro it on my local machine with the information you provided. (Comment added from ex-user id:gtiwari)
Comment by Adobe D.
18594 | August 08, 2012 01:07:53 AM GMT
Tried reproducing with frameset and 3 cfm pages. Not reproducible. Could you please share a reproducible sample of the issue. (Comment added from ex-user id:gtiwari)
Comment by Adobe D.
18595 | August 08, 2012 01:20:10 AM GMT
The problem is erratic. It occurs at most, one percent of the time. As a consequence, it is hard to reproduce. I managed to get it to happen with debugging turned on and have attached the "view source" contents from the web browser, for a request that was executed incorrectly. It is in HTML format of course.
Comment by External U.
18596 | August 08, 2012 11:53:31 AM GMT
If you examine the attachment (cust_inv_payment_2.txt) you will see that the debugging information indicates that the requested template was "cust_inv_payment_2.cfm". The CGI variable SCRIPT_NAME indicates the same thing. But that is not what actually was executed. What was actually executed was "cust_inv_payment_1.cfm" as you can verify by looking at the Execution Time table. The Execution Time table doesn't show an entry for the requested template at all, but rather for a different one that was requested and executed correctly immediately before (or at the same time). The web server log shows: 67.42.30.176 - - [08/Aug/2012:11:04:58 -0600] "GET /app/builder/cust_inv_payment.cfm?mode=job&init=1&vcfid=5321 HTTP/1.1" 200 3429 67.42.30.176 - - [08/Aug/2012:11:04:58 -0600] "GET /app/builder/cust_inv_payment_1.cfm?return=&p_frame=right&mode=job HTTP/1.1" 200 4676 67.42.30.176 - - [08/Aug/2012:11:04:58 -0600] "GET /app/builder/cust_inv_payment_2.cfm?return=&p_frame=right&mode=job HTTP/1.1" 200 4615 However cust_inv_payment_1.cfm was executed twice, instead of cust_inv_payment_1.cfm followed by (or in parallel with) cust_inv_payment_2.cfm, as requested.
Comment by External U.
18597 | August 08, 2012 12:09:09 PM GMT
The "debug_info.txt" attachment has the debugging information converted into text format, so it is easier to read.
Comment by External U.
18598 | August 08, 2012 12:14:35 PM GMT
this has been fixed in Update 4
Comment by Asha K.
18599 | November 07, 2012 05:35:19 AM GMT