Gravatar for erkan.yildiz@avanade.com

Question by erkany, Sep 8, 2015 1:30 PM

Rebuild index stuck at "Queued"

Hi,

I kicked a rebuild index from Sitecore control panel -> indexing manager.

The wizard shows up but gets stuck at :

"IndexUpdateIndexname=CoveoEUcdperf_index Queued" message and nothing happens from there

I don't see anything in Sitecore logs

How can I troubleshoot this issue

1 Reply
Gravatar for jflheureux@coveo.com

Answer by Jean-François L'Heureux, Sep 8, 2015 2:32 PM

  1. It may be a case of Sitecore endless index rebuild (see Endless Coveo Index Rebuild).
  2. Is your Coveo Diagnostic Page shows all components as green?
  3. Is your RabbitMQ service running?
  4. Is your RabbitMQ management tool accessible?
  5. Is there a "CoveoEUcdperf_index - [MachineName]-[SitecoreInstanceName]" queue in RabbitMQ?
  6. Are there any stuck messages in this RabbitMQ queue?
  7. Is your Coveo Enterprise Search (CES) service running?
  8. Is there a "CoveoEUcdperf_index - [MachineName]-[SitecoreInstanceName]" source in CES?
  9. Is your Coveo Admin Service service running?
  10. Is your Coveo.SearchProvider.config file correctly configured with the right <AdminServiceUri>, <Username>, <Password>, <QueueUri>, <QueueUsername> and <QueuePassword>?
  11. Do your "CoveoEUcdperf_index" index have the same type, param and crawler type as the other Coveo indexes?
  12. Do your "CoveoEUcdperf_index" index have a strategy?
Gravatar for erkan.yildiz@avanade.com

Comment by erkany, Sep 8, 2015 5:09 PM

Hi - everything seems ok other than Rabbit MQ queue shows 1000 unacknowledged messages. Any idea what might be the cause?

Gravatar for jflheureux@coveo.com

Comment by Jean-François L'Heureux, Sep 8, 2015 5:18 PM

Your CES source for that Sitecore index had an error, lost its connection to RabbitMQ and stopped consuming messages. You can reconnect it by:

  1. Opening the CES Administration Tool in the "Sources and Collections" section (Index > Sources and Collections).
  2. Clicking on your "CoveoEUcdperf_index - [MachineName]-[SitecoreInstanceName]" source to go to its status page.
  3. Clicking on the "Disable Live Monitoring" button (see Toggling Live Monitoring for a Source).
  4. Clicking on the "Enable Live Monitoring" button to reconnect the source to RabbitMQ and start consuming the stuck messages.
Gravatar for erkan.yildiz@avanade.com

Comment by erkany, Sep 8, 2015 5:29 PM

Thank you - Coveo started consuming messages back. This problem happened after indexing 40998 items. Do you know what might be the cause? This is a production environment and we would not want to babysit RabbitMQ every time we rebuild the index

Gravatar for jflheureux@coveo.com

Comment by Jean-François L'Heureux, Sep 8, 2015 5:37 PM

You can scan the CES system and index logs to look for any error and warning during that time. Report them here if you don't understand them.

It may be caused by a lot of factors including:

  • Not enough free disk space for RabbitMQ or Coveo index.
  • Crash of RabbitMQ service.

It may be important to note that the June 2014, October 2014 and November 2014 releases or Coveo for Sitecore included fixes for known bugs that had this exact behavior (messages stuck in queue). If you are using a release older than the November 2014 release, I highly suggest you to upgrade to the latest release.

Ask a question