Query results are erratic right after deployment during snapshot fetch

Users
1

Config: 1 zero 3 servers
Right after bulk load and bringing up the servers, queries are served with correct results intermittently, most queries fail with error like Invalid request, please try again. During this time I see that the servers are exchanging snapshots amongst themselves. The logs read
predicate.go:52: Getting SNAPSHOT: Time elapsed: 05s, bytes written: 110 MB, bytes/sec 22092938
Once this Getting snapshot is done across all the servers, the queries are served just fine.

  1. Is there any way to know when the Getting snapshot finishes, so we know that the server is ready to serve the requests properly?
    or
  2. Is there a way to disable the exchange of snapshots between the servers, since all the snapshots are identical (we are bringing up 3 servers with identical data)
2

this is a bug. you can refer this topic : Dgraph server log - #2 by mrjn

3

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.