-
- Downloads
Connection health monitoring
- Simplify node peers handling. - Remove connection reference counting. - Add connection health monitoring. - Use monitoring output to only send back a healthy connection when asked to get a server or leader for a group. Squashed commit of the following: commit bb1f68a044d51f3f61d3d16b5a6ff7c6656e97e1 Author: Manish R Jain <manish@dgraph.io> Date: Thu Sep 14 17:42:25 2017 +1000 No need to send export call to dgraphzero, it won't handle export anyway. commit 0106decd948fc5f28242d9b098ca9cccbcbe31fc Author: Manish R Jain <manish@dgraph.io> Date: Thu Sep 14 17:02:09 2017 +1000 Tested that the connection switch automatically happens in favor of healthy one. commit f4bfdb40ecf5f17a1cebc0f9d3f5249433d4c79c Author: Manish R Jain <manish@dgraph.io> Date: Thu Sep 14 16:34:49 2017 +1000 Add connection health checking. Also before returning connections to servers in a group, check health before returning. commit 70ebcc44a5ab8731e10a027689a084cbe78885a2 Author: Manish R Jain <manish@dgraph.io> Date: Thu Sep 14 16:17:13 2017 +1000 Remove refcounting. And start monitoring health of connection.
Showing
- cmd/dgraphzero/raft.go 0 additions, 1 deletioncmd/dgraphzero/raft.go
- cmd/dgraphzero/zero.go 1 addition, 2 deletionscmd/dgraphzero/zero.go
- conn/node.go 19 additions, 25 deletionsconn/node.go
- conn/pool.go 45 additions, 133 deletionsconn/pool.go
- worker/draft.go 7 additions, 12 deletionsworker/draft.go
- worker/export.go 6 additions, 37 deletionsworker/export.go
- worker/groups.go 18 additions, 29 deletionsworker/groups.go
- worker/mutation.go 9 additions, 13 deletionsworker/mutation.go
- worker/schema.go 3 additions, 5 deletionsworker/schema.go
- worker/task.go 0 additions, 1 deletionworker/task.go
Loading
Please register or sign in to comment