-
Type: Bug
-
Resolution: Fixed
-
Priority: Major - P3
-
Affects Version/s: 5.0.0, 6.0.0, 7.0.0, 8.0.0-rc0
-
Component/s: Sharding
-
Catalog and Routing
-
Fully Compatible
-
ALL
-
v8.0, v7.0, v6.0, v5.0
-
CAR Team 2024-07-08, CAR Team 2024-07-22, Cluster Scalability 2024-6-10, CAR Team 2024-08-05, CAR Team 2024-08-19, CAR Team 2024-09-02
-
200
-
3
ShardRegistry does a periodic _lookup of cache every 30 seconds by default. We currently force a refresh pessimistically and in doing so block all threads that need shard registry data for targeting shards over the network. This can be bad for user threads which get blocked until the refresh completes, particularly for colocated/non-dedicated config servers when there are lagging secondaries and the reads go to them.
- is duplicated by
-
SERVER-90841 ShardRegistry periodic refresh forces others to block on refresh as well
- Closed
- is related to
-
SERVER-90841 ShardRegistry periodic refresh forces others to block on refresh as well
- Closed
- related to
-
SERVER-93101 Fix ShardRegistry::Time definition
- Closed
-
SERVER-93700 Explore using RSM monitor pings to update the topologyTime for ShardRegistry
- Backlog