-
Type:
Task
-
Resolution: Won't Do
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: Testing Infrastructure
-
Service Arch
-
Service Arch 2018-10-08, Service Arch 2018-10-22, Service Arch 2018-11-05, Service Arch 2018-11-19, Service Arch 2018-12-03
-
None
-
3
-
None
-
None
-
None
-
None
-
None
-
None
-
None
What happens when we bounce / bring on a new mongos
Test:
- Set up a 1 shard cluster
- Set up a workload of N sleeps per second via a mongos
- Stage these out so that the workload practically requires M connections to execute (due to the sleeps)
- Measure steady state perf
- Try stuff:
- Set up a second mongos
- Point an equivalent workload against it
- Bounce the mongos
- Time to reversion to mean
- Set up a second mongos
- Variants:
- Is it worse if there is a very large routing table
- Many shards
- Many chunks
- Is it worse if there is a very large routing table