No connection exist

Users
1

When I query in UI, it returns an error No connection exist

dgraph-zero:
2017/12/23 20:57:19 oracle.go:375: No healthy connection found to leader of group 52
2017/12/23 20:57:19 pool.go:167: Echo error from localhost:7101. Err: rpc error: code = Unavailable desc = all SubConns are in TransientFailure

It reports that it lost connection to the dgraph server with port 7101, but that dgraph server does not report any error.

last line of dgraph server (port 7101)
2017/12/23 20:20:54 lru.go:79: LRU cache max size is being set to 50 MB

2

Group 52? That’s a very high group number. I reckon what might have happened is that during your experimentation, you didn’t provide an --idx to Dgraph and deleted its w or p directories, due to which Zero assigned it a new group every time.

You should clear out all the dgraph and zero directories and start the cluster, this time providing a consistent unique index via the flag.