Index stuck in bootstrap, how to deal with that?

Couchbase Server
#1

Hi,

One of my indexes is in bootstrap mode and couldnot be brought to ready state… The questions I have are:

  1. How to make this index in a working state?
  2. If 1 doesn’t work, would drop and re-create the index work? ( how do you deal with the timeout on the drop index operation?)

thanks
edward

#2

Hi, did you try running the BUILD INDEX statement as explained here: http://developer.couchbase.com/documentation/server/4.5/n1ql/n1ql-language-reference/build-index.html ?

#3

I issued the query SELECT * FROM system:indexes and verified that the index were removed. However, I could still see it in the ‘Indexes’ tab from the Administrator web portal. And it complained about “Warning: Cannot communicate with indexer process. Information on indexes may be stale. Will retry.” under Global Indexes, and the index supposed to be removed is still in bootstrap status.

Anyway, I ran a query ‘create primary index’ to recreate that index. And the “SELECT * FROM system:indexes” returned ‘building’ state. It has been over 24 hours. I wonder whether there is a good way to check the percentage of the indexing building?

thanks
edward

#4

I am seeing the exact same issue as this (4.5 Enterprise). I did manage to delete the index and restart the server and when I tried to recreate the index again, it gets “stuck” and I am seeing the “Warning: Cannot communicate with indexer process. Information on indexes may be stale. Will retry.” message without even seeing the index in the list.

Seems like somehow the index process is screwed up somewhere, but I am not sure where to start fixing it (except by uninstall and reinstall (which is a pain but is possible - it’s a test environment). But that won’t really help me avoid the problem in the future.

Any help would be appreciated.

Scott

#5

Any solution for this, I got same issue for 6.5.1
No error from indexer.log, in “htop”, can see 2 thread of indexer always 100%

Thanks