Gets this error "Error communicating to server! timeout" every time I try to look at my pipeline. It only works like 1 time on 10 to 20 tries.

      Jenkins version: 1.599
      Delivery Pipeline Plugin: 0.8.9

          [JENKINS-27354] Error communicating to server! timeout

          I don't get a stacktrace, it's not printed. The only thing printed is "Error communicating to server! timeout"

          We have 3 piplines and like 4 stages on each.

          Anders Nyström added a comment - I don't get a stacktrace, it's not printed. The only thing printed is "Error communicating to server! timeout" We have 3 piplines and like 4 stages on each.

          Jouni Rajala added a comment -

          I have same issue, but noticed it has something to do with my slave confuguration. Jobs that are run on master server are shown in diagram, but anything run on slave machines causes 'error communicating to server! timeout'

          Configuration consist windows master and multiple windows slave machines.

          Jouni Rajala added a comment - I have same issue, but noticed it has something to do with my slave confuguration. Jobs that are run on master server are shown in diagram, but anything run on slave machines causes 'error communicating to server! timeout' Configuration consist windows master and multiple windows slave machines.

          Alex Ehlke added a comment -

          We also run into this. I'm not sure why, but it may be because we also have slave jobs in each pipeline.

          Alex Ehlke added a comment - We also run into this. I'm not sure why, but it may be because we also have slave jobs in each pipeline.

          Alex Ehlke added a comment -

          We have a fluctuating # of pipelines, generally around 50 at a time (and each pipeline has some dozen instances or so).

          Alex Ehlke added a comment - We have a fluctuating # of pipelines, generally around 50 at a time (and each pipeline has some dozen instances or so).

          Alex Ehlke added a comment - - edited

          Just reproduced, it's the call to /api/json (e.g. in our case, https://foo.example/job/Branches/view/feature-box_polling/api/json ) which times out with this nginx message: "504 Gateway Time-out"

          I also removed all but one pipeline, and set it to only have 2 instances, and I'm having the same issue. So it seems linked to the potential slave jobs cause.

          Alex Ehlke added a comment - - edited Just reproduced, it's the call to /api/json (e.g. in our case, https://foo.example/job/Branches/view/feature-box_polling/api/json ) which times out with this nginx message: "504 Gateway Time-out" I also removed all but one pipeline, and set it to only have 2 instances, and I'm having the same issue. So it seems linked to the potential slave jobs cause.

          Alex Ehlke added a comment -

          Please let me know if any other info would be useful, I'm eager to assist getting this fixed however I can Thanks.

          Alex Ehlke added a comment - Please let me know if any other info would be useful, I'm eager to assist getting this fixed however I can Thanks.

          Patrik Boström added a comment - - edited

          Need some information to understand why the call to https://foo.example/job/Branches/view/feature-box_polling/api/json takes that long. Threaddumps would be helpful during the call that timeouts.

          Patrik Boström added a comment - - edited Need some information to understand why the call to https://foo.example/job/Branches/view/feature-box_polling/api/json takes that long. Threaddumps would be helpful during the call that timeouts.

          Alex Ehlke added a comment -

          @Patrik, here's a thread dump of master and slave (I believe slave was building a job at the moment), in another couple tabs I had the delivery pipeline view open and these GETs can be seen in the dump: https://dl.dropboxusercontent.com/u/2112716/Thread%20dump%20%5BJenkins%5D%20delivery%20pipeline%20timeout.html

          Alex Ehlke added a comment - @Patrik, here's a thread dump of master and slave (I believe slave was building a job at the moment), in another couple tabs I had the delivery pipeline view open and these GETs can be seen in the dump: https://dl.dropboxusercontent.com/u/2112716/Thread%20dump%20%5BJenkins%5D%20delivery%20pipeline%20timeout.html

          Alex Ehlke added a comment -

          I found the issue: I have a job called "Job-DSL" which is a pipeline step, it runs the Netflix Job DSL very frequently. I wasn't discarding old builds, so thousands piled up. Once I cleared the builds directory, the delivery pipeline started working (and snappily) again.

          Alex Ehlke added a comment - I found the issue: I have a job called "Job-DSL" which is a pipeline step, it runs the Netflix Job DSL very frequently. I wasn't discarding old builds, so thousands piled up. Once I cleared the builds directory, the delivery pipeline started working (and snappily) again.

          Tommy Tynjä added a comment -

          Marking as resolved since it seems to be caused by job configurations preserving a lot of history. A workaround is to preserve less history.

          Tommy Tynjä added a comment - Marking as resolved since it seems to be caused by job configurations preserving a lot of history. A workaround is to preserve less history.

            patbos Patrik Boström
            midikem Anders Nyström
            Votes:
            1 Vote for this issue
            Watchers:
            6 Start watching this issue

              Created:
              Updated:
              Resolved: