Check health timeout and period is too short, and will lead to the following error in OLAP engine:
2021-12-08 05:18:47,334 WARN org.tikv.common.region.StoreHealthyChecker [] - store [tikv_host:tikv_port] is not reachable
2021-12-08 05:19:01,130 WARN org.tikv.common.region.AbstractRegionStoreClient [] - forward request to store [tikv_host:tikv_port] for region[174126738]
2021-12-08 05:19:01,894 WARN org.tikv.common.region.StoreHealthyChecker [] - store [tikv_host:tikv_port] is not reachable
2021-12-08 05:19:01,933 WARN org.tikv.common.region.StoreHealthyChecker [] - store [tikv_host:tikv_port] recovers to be reachable and canforward
2021-12-08 05:19:05,490 WARN org.tikv.common.region.StoreHealthyChecker [] - store [tikv_host:tikv_port] is not reachable
2021-12-08 05:19:05,604 WARN org.tikv.common.region.StoreHealthyChecker [] - store [tikv_host:tikv_port] recovers to be reachable and canforward
In the big data scenario, we don't care about latency, so check health timeout and period should be configurable.
Check health timeout and period is too short, and will lead to the following error in OLAP engine:
In the big data scenario, we don't care about latency, so check health timeout and period should be configurable.