Correlation Parameter Problems (Error 35049)

[Jeremy K]

Can anybody share their experiences with Error Code -35049 which is a problem with the Correlation Parameter?

 

I'm testing a web reporting application using a script created in Vugen v11.52 with the 'Web - HTTP/HTML' protocol. I can run the script from Vugen and it works fine. When I execute the script against the app with multiple virtual users, I consistently get this error and the test fails:  Action_ML.c(1142): Error -35049: No match found for the requested parameter "CorrelationParameter_1". Check whether the requested regular expression exists in the response data, Snapshot Info [MSH 1 142].

 

I've run the test numerous times, I would say it runs successfully 1 out of every 10 attempts and it seems like it will run OK with just one user, but the multiple virtual users seems to result in this error.  I've re-created the script several times and I've modified the script based on a smattering of advice I've seen in this group and others.

 

I grabbed this log from one attempt:

Start auto log messages stack - Iteration 1. [MsgId: MMSG-10545]
Action_ML.c(1066): web_custom_request("validate_2") was successful, 2294 body bytes, 865 header bytes, 12 chunking overhead bytes [MsgId: MMSG-26385]
Action_ML.c(1130): web_reg_save_param_regexp started [MsgId: MMSG-26355]
Action_ML.c(1130): Registering web_reg_save_param_regexp was successful [MsgId: MMSG-26390]
Action_ML.c(1138): Notify: Transaction "Create_ML_Report" started. [MsgId: MMSG-16999]
Action_ML.c(1140): lr_think_time: 8.00 seconds (recorded think time was 16.00 seconds). [MsgId: MMSG-15947]
Action_ML.c(1142): web_custom_request("cache") started [MsgId: MMSG-26355]
Action_ML.c(1142): Redirecting "[url]" (redirection depth is 0) [MsgId: MMSG-26694]
Action_ML.c(1142): To location "[url]"  [MsgId: MMSG-26693]
Action_ML.c(1142): Error -35049: No match found for the requested parameter "CorrelationParameter_1". Check whether the requested regular expression exists in the response data, Snapshot Info [MSH 1 142]   [MsgId: MERR-35049]
End auto log messages stack. [MsgId: MMSG-10544]

 

[James Pulley]

Several observations:

• The page you expect may not be the page you are receiving back for processing.   The page may be an HTTP 200 page, but it may simply say something like "I have an error" in friendly terms.   Make sure you are checking content for each and every request and then branching your code when you receive a page which is not appropriate
• The use of parameters in the pattern of <param name>_<some number> is typical of correlation rule driven correlation parameters.  In such a case, unless you have written the rule for optimal xpath of other conventions, the parameters of the correlation function may not be optimal for all page and data conditions.   In this case you will need to use your knowledge of manual correlation to go in and tune the parameter conventions to ensure a higher rate of success
• I notice from the log above you have your think time inside of your timing record. Not considered a good practice. This practice is associated with newer users of the tool so you should expect to have a mentor available to you onsite while you are coming up to speed.  They should also be able to assist in diagnosing the correlation function options.

[Jiikka]

Hey Jeremy,

I have had a similar issue, good chance it could it be the same. 

Check you cache setting

   i) In your Run time settings, under Browser Emulation see if you have checked "clear cache on each iteration"

or 

   ii) Use the function in your script web_cache_cleanup 

1st user will pass every time and without the cache being cleaned all the other users will fail.

regards,

Jikka.

[jinal...@gmail.com]

You can find the solution if you check the logs. 

On Tuesday, December 9, 2014 12:29:01 PM UTC-5, Jeremy K wrote:

[Bug Man]

Jeremy K,
You said "it seems like it will run OK with just one user,"

is it using session id for each user? if each user use different names in the script like (abc, cnc, bnc...) For example in the script abc uses101 , bnc uses 309, cnc uses 486 like.
I have experienced the same problem and corellate 101 value that solvs the issue.
You may correlate abc but forgod to correlate 101. or you may forgot correlate both. if that is the case correlate both then run. increase the think time.

Try.

On Tuesday, December 9, 2014 12:29:01 PM UTC-5, Jeremy K wrote:

[Pradhyumna Gupta]

Is this the first request in your script?
Try running the script with different data and see if it works.

If this is not the 1st request then make sure you have handled all dynamic values in all the previous requests. Put a check to ensure each request is giving you the expected response.

--
You received this message because you are subscribed to the Google Groups "LoadRunner" group.
To unsubscribe from this group and stop receiving emails from it, send an email to LR-LoadRunne...@googlegroups.com.
To post to this group, send email to LR-Loa...@googlegroups.com.
Visit this group at http://groups.google.com/group/LR-LoadRunner.
For more options, visit https://groups.google.com/d/optout.

[Jeremy K]

There is not a session id, but there are unique ids generated upon certain actions in the application.  I'll try your advice with that.  Thank you.

[Jeremy K]

It's not the first request in the script.  I'll look again at my dynamic values, thanks!