Refresh on index hangs

Elastic Stack Elasticsearch
1

Hi All,
Running refresh on an index sometimes hangs forever and sometimes works (from python,
using elasticsearch module):

r = requests.post(es_url+'/' + myindex + '/_refresh?pretty', auth=(username, password))
print(r.text)

The index is tiny. Couple small documents were inserted into it before running refresh.
Restarting Elastic Search fixes the problem for a while but then it starts again.
I am using Elastic Search version 6.2.1. There are 7 nodes in the cluster.
Thank you,
Igor

2

When you say "forever" how long are you waiting? Is the cluster otherwise responsive? What kind of load does it see?

Can you reproduce with curl? Would be good to rule out something client-side.

Can you look at GET _nodes/hot_threads?threads=9999 to see if there's a thread stuck doing the refresh somewhere, and also look at GET _tasks to see if the refresh task is still pending?

Is there any suggestion of something going wrong in the server logs?

There were quite a few changes to the replication logic in 6.3. Can you reproduce this on a more recent version?

3

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.