Coveo indexing read only due to disk being full
I am trying to index our QA server, but I keep getting an error saying this:
Job started: IndexUpdateIndexName=sitecoremasterindex|#Exception: System.Reflection.TargetInvocationException: Exception has been thrown by the target of an invocation. ---> System.ServiceModel.FaultException`1[System.ServiceModel.ExceptionDetail]: The requested operation cannot be performed in read-only mode.
I keep resetting this to read and write mode, but it keeps switching. After looking into this more, I see that the hard drive is almost full. Looking at the files, I found that this directory alone is filling up the entire hard drive (RabbitMQ\RabbitMQ for Coveo\db\rabbit@WIN-9LCLLMABRM9-mnesia\msgstorepersistent).
All the files in there are .rdq files, would there be any issues if I just deleted these and re-run the indexing or how should I handle this?
The queues should empty by themselves during the indexing operation. RabbitMQ only acts as a buffer between Sitecore and Coveo so messages should not be kept in it. If it does, I would try to restart Live-monitoring on the Sitecore source (in the Coveo Administration Tools) while keeping an eye in the CES Console for any errors.
You can also change the location of the queue: https://developers.coveo.com/display/public/SC201409/Changing+the+Location+of+RabbitMQ+Queues
If there are no errors but the queues are stuck with messages in it, I would advise to open a support case.