Uploaded image for project: 'Couchbase Java Client'
  1. Couchbase Java Client
  2. JCBC-1809

WaitUntilReady against buckets failing on <6.5 clusters

    XMLWordPrintable

Details

    • Bug
    • Status: In Progress
    • Major
    • Resolution: Unresolved
    • 3.1.5, 3.2.1
    • None
    • None
    • 1

    Description

      There are consistent failures across the WaitUntilReadyTests for the tests running against 6.0 and 5.5 (also 7.0 - refer to comment), specifically these tests:

      • handlesCreatingBucketDuringWaitUntilReady
      • waitsForNewlyCreatedBucket

      The waitsForNewlyCreatedBucket test has the following error, on a drop bucket call:
      com.couchbase.client.core.error.FeatureNotAvailableException: The cluster does not support cluster-level queries (only Couchbase Server 6.5 and later) and no bucket is open. If an older Couchbase Server version is used, at least one bucket needs to be opened to dispatch this mgmt request!

      And the handlesCreatingBucketDuringWaitUntilReady test times out, with retry reasons: BUCKET_OPEN_IN_PROGRESS.

      My understanding is that the waitUntilReady should work against buckets only for <6.5 servers - at least, that is the behaviour in Gocb. So these tests should work.

      Attachments

        Issue Links

          No reviews matched the request. Check your Options in the drop-down menu of this sections header.

          Activity

            will.broadbelt Will Broadbelt added a comment - - edited

            Michael Nitschinger : I've noticed that the waitsForNewlyCreatedBucket test is also rarely failing against 7.0.0 (on Jenkins) due to a reactor overflow exception:

            waitsForNewlyCreatedBucket failure

            reactor.core.Exceptions$OverflowException: Could not emit tick 552 due to lack of requests (interval doesn't support small downstream requests that replenish slower than the ticks)
            at com.couchbase.client.java.AsyncUtils.block(AsyncUtils.java:51)
            at com.couchbase.client.java.Bucket.waitUntilReady(Bucket.java:232)
            at com.couchbase.client.java.WaitUntilReadyIntegrationTest.waitsForNewlyCreatedBucket(WaitUntilReadyIntegrationTest.java:135)
            at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
            at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
            at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
            at java.lang.reflect.Method.invoke(Method.java:498)
            at org.junit.platform.commons.util.ReflectionUtils.invokeMethod(ReflectionUtils.java:688)
            etc.

            Full error:

             http://sdk.jenkins.couchbase.com/view/Nightly/job/jvm/job/jvm-clients/job/couchbase-jvm-clients-scipted-build-pipeline/734/testReport/com.couchbase.client.java/WaitUntilReadyIntegrationTest/testing___Linux__cbdyncluster_7_0_stable__Oracle_JDK_8____waitsForNewlyCreatedBucket_1_/

            Looks like a similar issue to JVMCBC-967, suggesting that perhaps its another race condition happening. I've found that it doesn't easily reproduce locally.

            This is also likely to be related to the recently filed issues that are affecting TAF tests: JCBC-1806, JCBC-1811.

            will.broadbelt Will Broadbelt added a comment - - edited Michael Nitschinger : I've noticed that the waitsForNewlyCreatedBucket test is also rarely failing against 7.0.0 (on Jenkins) due to a reactor overflow exception: waitsForNewlyCreatedBucket failure reactor.core.Exceptions$OverflowException: Could not emit tick 552 due to lack of requests (interval doesn't support small downstream requests that replenish slower than the ticks) at com.couchbase.client.java.AsyncUtils.block(AsyncUtils.java:51) at com.couchbase.client.java.Bucket.waitUntilReady(Bucket.java:232) at com.couchbase.client.java.WaitUntilReadyIntegrationTest.waitsForNewlyCreatedBucket(WaitUntilReadyIntegrationTest.java:135) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.lang.reflect.Method.invoke(Method.java:498) at org.junit.platform.commons.util.ReflectionUtils.invokeMethod(ReflectionUtils.java:688) etc. Full error:   http://sdk.jenkins.couchbase.com/view/Nightly/job/jvm/job/jvm-clients/job/couchbase-jvm-clients-scipted-build-pipeline/734/testReport/com.couchbase.client.java/WaitUntilReadyIntegrationTest/testing___Linux__cbdyncluster_7_0_stable__Oracle_JDK_8____waitsForNewlyCreatedBucket_1_/ Looks like a similar issue to JVMCBC-967 , suggesting that perhaps its another race condition happening. I've found that it doesn't easily reproduce locally. This is also likely to be related to the recently filed issues that are affecting TAF tests: JCBC-1806 , JCBC-1811 .
            daschl Michael Nitschinger added a comment - Change for overflow exception is in review: http://review.couchbase.org/c/couchbase-jvm-clients/+/155287

            Can you rerun with latest master (all changes merged) and see if the issue still shows up?

            daschl Michael Nitschinger added a comment - Can you rerun with latest master (all changes merged) and see if the issue still shows up?

            The Overflow exception does not show up in my testing anymore so am confident it is fixed. Though the titular issue with <6.5 buckets is still present (probably should've been separate issues...).

            will.broadbelt Will Broadbelt added a comment - The Overflow exception does not show up in my testing anymore so am confident it is fixed. Though the titular issue with <6.5 buckets is still present (probably should've been separate issues...).

            Michael Nitschinger - Any updates on this issue? WaitUntilReady on buckets for clusters <6.5 is still failing regularly in the combination pipeline.

            will.broadbelt Will Broadbelt added a comment - Michael Nitschinger  - Any updates on this issue? WaitUntilReady on buckets for clusters <6.5 is still failing regularly in the combination pipeline.

            People

              daschl Michael Nitschinger
              will.broadbelt Will Broadbelt
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated:

                Gerrit Reviews

                  There are no open Gerrit changes

                  PagerDuty