Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

See, there are nothing called intermittent issue. Just wrong way of doing stuff :) On Fri, Sep 11, 2015 at 3:36 PM, Niranjan Karunanandham wrote: > Hi Shankar, > > I went through the log and noticed that in the ServerRestartTestCase, it > sends a command to restart the server

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi Shankar, I went through the log and noticed that in the ServerRestartTestCase, it sends a command to restart the server and waits for 15 seconds and proceeds. If in case the server takes longer time then there is an issue. Therefore I added the check to wait for the http port to be closed (or

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Even waiting for 30 seconds might not work. Can we loop with a sleep and see whether the process is not there and proceed? Port will be closed immediately, but the process might still run AFAIK. On Fri, Sep 11, 2015 at 3:41 PM, Selvaratnam Uthaiyashankar < shan...@wso2.com> wrote: > See, there

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi Shankar, We wait for 5 min(by looping till 5 min) to let the server port open. if the server failed to open the port within 5 min, We failed the test case or skip all the tests. However 30 sec may be not enough for server to close the port when restart, if it fail, Test case will be failed and

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi Nuwan, My comment is, don't assume a particular duration where it will happen. I agree we have to have a maximum, and beyond that we have to fail the test case. If we are doing that, then fine. Just wanted to make sure we are not assuming if we wait for n minutes, anyway the port will be

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi Shankar, In the code currently we are looping to check if the port is close. If the port does not close within 30 seconds then there should be an issue (please correct me if am wrong). Then we wait for 15 seconds and then check if the port is open (using the waitForPort provided by the

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Yes. We have to check whether the port is closed by looping within a defined max time (30 sec) just after *restart* is called. then we know It is shutdown and going to restart. Then we have to wait for port to open. Note : 30 sec is depends on the server load at that time. We can increase the

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

sorry I misunderstood your comment previously.. Yes, what you have done is correct, but my thinking is, what will happen if the port is not closed within 30 seconds? Do we fail the test case? Same for the check on port open. If it didn't open within 15 seconds, do we fail the test case? On Fri,

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi Supun, I will look into this. Just now triggered a build in openstack. Regards, Nira On Fri, Sep 11, 2015 at 8:33 AM, Supun Malinga wrote: > org.wso2.carbon.integration.tests.integration.ServerShutdownTestCase.testGracefulServerShutdown > test method fails saying below, >

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi all, I was able to build the Kernel 4.4.x branch successfully. As Supun mentioned this looks like an intermittent issue. I will look into this further and update this thread. Regards, Nira On Fri, Sep 11, 2015 at 9:28 AM, Niranjan Karunanandham wrote: > Hi Supun, > > I

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Hi Shankar, It seems to be another server was accessing the h2 database. I am looking into the CarbonTestServerManager class which is used to start the servers. Regards, Nira On Fri, Sep 11, 2015 at 10:36 AM, Selvaratnam Uthaiyashankar < shan...@wso2.com> wrote: > Folks, intermittent issues

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

Folks, intermittent issues are the serious problem. We can't just leave it with intermittent issues. Please track and see what was the cause? database locked by another process means, two servers are running at the same time? On Fri, Sep 11, 2015 at 10:14 AM, Niranjan Karunanandham

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

org.wso2.carbon.integration.tests.integration.ServerShutdownTestCase.testGracefulServerShutdown test method fails saying below, Aruna/ Niranjan can one of you have a look ? This is 4.4.x branch. fyi, I built 4.4.x last night with all tests and build passed for me. May be this is intermittent.

Re: [Dev] Jenkins build became unstable: carbon4-kernel_4.4.x #48

On Fri, Sep 11, 2015 at 10:38 AM, Niranjan Karunanandham wrote: > Hi Shankar, > > It seems to be another server was accessing the h2 database. > How it is possible? It can't be intermittent.. May be we are shutting down / killing the server in one test case and without