Uploads Stay in Queue - Fluid Topics - 3.8

Fluid Topics Troubleshooting Guide

Product
Fluid Topics
FT_Version
3.8
Platform
RHEL
Category
Reference Guide
language
English

Issue

On the Publish interface, uploads stay in queue even after a long period of time.

Impact(s) on users

  • This issue impairs the display of the Publish interface for KHUB Admin.
  • This issue causes no impact to end-users.

Possible root cause(s)

This issue might occur when:

  • A PaF ends in error and does not warn the server that it is finished.

Solution(s)

Solution 1

This solution is only for clients with an on-premises license.

Connect to the Front Mongo cluster and delete the running jobs.

In the case of a remote connection to the cluster:

  1. Go to the front-end SSH server.

    Example:

    ssh fluid-front-end1

  2. Connect remotely to the SSH server where the Mongo databases are located.

    Example:

    mongo -u $USER-p --authenticationDatabase admin fluid-dbfront-end...

    • To know the name of the MongoDB primary node, run the following command:

      rs.isMaster()

      The following lines show a result for this command where the primary node is fluid-dbfront-end3:27017

      {
      "hosts" : [
      "fluid-dbfront-end1:27017",
      "fluid-dbfront-end2:27017",
      "fluid-dbfront-end3:27017"
      ],
      ...
      "primary" : "fluid-dbfront-end3:27017",
      ...
      }

      OR

    • Run the following command:

      rs.status()

  3. Run the following commands to list the databases:

    mongo> show dbs

  4. Run the following command to select a database:

    mongo> use fluid-topics-$SERVICE_ID-$SERVICE_STATUS

  5. Run the following command to find all running uploads:

    mongo> db.publish_reports.find({'status': 'RUNNING'});

  6. Run the following command to remove all running uploads:

    mongo> db.publish_reports.remove({'status': 'RUNNING'}, {'multi': true});

If this procedure returns a "Not available currently" error message, please contact the Antidot Support.

There is no need to repeat this procedure on other Mongo databases. The modifications are automatically replicated to the other cluster nodes.

In the case of a direct connection to the cluster:

  1. Run the following command:

    *ssh fluid-dbfront-end
    *mongo

  2. Run the following commands to list the databases:

    mongo> show dbs

  3. Run the following command to select a database:

    mongo> use fluid-topics-$SERVICE_ID-$SERVICE_STATUS

  4. Run the following command to find all running uploads:

    mongo> db.publish_reports.find({'status': 'RUNNING'});

  5. Run the following command to remove all running uploads:

    mongo> db.publish_reports.remove({'status': 'RUNNING'}, {'multi': true});

If this procedure returns a "Not available currently" error message, please contact the Antidot Support.

There is no need to repeat this procedure on other Mongo databases. The modifications are automatically replicated to the other cluster nodes.

Check that the vhost configuration has been properly set on the server(s) hosting the front environment and the indexing environment. If needed, modify the vhost.conf.json Configuration File.