At time of NFT testing , mq stubs are automatically going to ended state and I am unable to restart the services in console .
The responses are getting timed out . could you please help me how to fix the timeout issue from stubs
As you are using Lisa 7.5.2, i doubt you will see Scope as it exists in IBM Native step which is not present in Lisa 7.5.2
Can you please try :
I hope this should solve you issue.
Questions that might help us help you are...
Which version of DevTest are you using?
When you examine the service in the console application, do you see a Red Dot next to your service or does your service indicate that it has erred out and shut itself down?
What messages do you see in the log files when the service shuts down and when you attempt to restart the service from the console?
Can you describe any customizations that exist in the VSM such as not sending responses for certain transactions, assertion branches in the VSM to Error, Generate Warning, etc.?
If you redeploy the service using Workstation or Console, does the service start correctly?
Please find answers below.
1. We are using CA Lisa 7.5.2 version.
2. No Joel, there is not any error or red warning or dot. Simply it is going to end state after 10 K or 20 K payments/messages injection. At front end , they are seeing time out as they are suspecting Lisa is sending response very late. But we have checked in service logs , there was not any delay in response.
3. We are seeing some service got overloaded error in logs during restart the service , also during when service is going to end state automatically, overload message is there in logs. But dont know how to resolve this and what actions to take.
As part of actions, we have removed all old logs file which were in MB , also removed or stop all other services except these to see if any good result but no luck.
4. There is not any custom code, filter, assertion in VSM. But we are giving capacity as 10 during deployment of VSM.
5. If we are deploying VSM very first then it is getting deployed fine but after some runs when we are trying to remove or deploy service in Console, some time we are some errror related to VSE.
Please check and assist.
Since you are running a capacity of 10, I am guessing you are running a performance VSE, correct?
Are you using the OOTB Derby database or have you connected LISA to an enterprise DBMS?
Is it possible to scrub any customer specific information and place some of the log messages you see in this post?
Your questions also lead me to believe that you are using JMS. If so, which JMS provider is being used and what do you have defined as the scope of your Assets?
- Step: An open asset instance is cached and reused for the duration of the step in which it is accessed. When the step is finished, the asset instance is closed. - Model: An open asset instance is cached and reused for the duration of the model execution in which it is accessed. When the model execution is finished, the asset instance is closed.
- Staged: An open asset instance is cached and reused by all instances of the model in which it is accessed. When all instances of the staged model are finished, the asset instance is closed.
Yes I am using Performance VSE. And I have checked it is connected to some SQL data base. I am using IBM MQ data protocol and I dont know where to check scop of asset, can you please tell me, so that I can check and confirm you?
Is it in VSM or VSE..please sahre me screen shot ?
In 7.x the connection assets are defined a little differently to that in current versions. There should be a "shared connections" option that you can enable.
As Dave says, they are a little different in 7.x. Open the project and look at where your connections are defined. The below are NOT IBM MQ connection assets, but use PRO mode and look for a dialog that looks something like this:
Did Dave's and Joel's responses help you out with this issue?
I am not able to find this option. And at one run service is running fine but another run on second day it is stopping in between test. I am doubting it can be performance VSE issue and we are giving 30 as capacity, is this capacity is fine or we are giving big number?
Thanks Sumeet. We have tried some of them also one more thing, we increased ulimit (-n) value which was very low. So now we ae not facing this kind of issue.
Thanks for you reply.
Retrieving data ...