Uploaded image for project: 'Nuxeo Platform'
  1. Nuxeo Platform
  2. NXP-27636

Some Gatling simulations timeout with AWS ALB or ELB

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Resolved
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: None
    • Fix Version/s: 11.1, 2021.0
    • Component/s: Core
    • Release Notes Description:
      Hide

      Nuxeo's Gattling simulations use long request to wait for long asynchronous tasks like indexing.
      When using a NAT router this inactive connections can be dropped resulting in request timeout.
      To prevent this TCP Keepalive needs to be enabled at OS level on the Gatilng node.
      In addition, it also requires to use Gatling > 3.2.1.

      Show
      Nuxeo's Gattling simulations use long request to wait for long asynchronous tasks like indexing. When using a NAT router this inactive connections can be dropped resulting in request timeout. To prevent this TCP Keepalive needs to be enabled at OS level on the Gatilng node. In addition, it also requires to use Gatling > 3.2.1.
    • Epic Link:
    • Tags:
    • Team:
      PLATFORM
    • Sprint:
      nxplatform 11.1.15, nxplatform 11.1.16, nxplatform 11.1.17
    • Story Points:
      2

      Description

      When the benchmark Gatling simulations are run though an Amazon load balancer to distribute the load on the Nuxeo Nodes,
      some long request timeouts while Nuxeo has returned the response without error.
      This has been seen with an AWS ALB or ELB (classical load balancer).
      This happens on Sim10MassImport Elasticsearch.WaitForIndexing and BulkUpdateDocuments in Bulk.WaitForAction.

      Both cases are long requests (more than 5min) that return a short response, the nuxeo access.log trace a correct response (200), the ELB also but the response is not received by Gatling which waits until the 1h timeout.

      This doesn't break the benchmark but makes it very slow and impossible to get metrics on async duration (indexation, bulk action).

        Attachments

          Issue Links

          depends on

          Improvement - An improvement or enhancement to an existing feature or task. VEND-44 Release Gatling with TCP Keepalive feature

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Resolved
          is related to

          Improvement - An improvement or enhancement to an existing feature or task. NXP-27977 Upgrade to Gatling > 3.2.1

          • Major - Major loss of function.
          • Resolved
          links to

          Web Link Gatling TCP Keepalive support #3790

            Activity

              People

              • Votes:
                0 Vote for this issue
                Watchers:
                4 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved:

                  Time Tracking

                  Estimated:
                  Original Estimate - 0 minutes
                  0m
                  Remaining:
                  Remaining Estimate - 0 minutes
                  0m
                  Logged:
                  Time Spent - 1 week, 1 day, 3 hours
                  1w 1d 3h