diff --git a/performance-tuning-methods.md b/performance-tuning-methods.md index 45c7d746f8d07..07bc41909e1ad 100644 --- a/performance-tuning-methods.md +++ b/performance-tuning-methods.md @@ -409,7 +409,7 @@ The TSO wait time is recorded as `TSO WAIT` and the network time of the TSO requ The indicators in this section correspond to the following three panels. - Avg TiDB KV Request Duration: The average latency of KV requests measured by TiDB -- Avg TiKV GRPC Duration: The average latency in processing gPRC messages in TiKV +- Avg TiKV GRPC Duration: The average latency in processing gRPC messages in TiKV - PD TSO Wait/RPC Duration: TiDB executor TSO wait time and network latency for TSO requests (RPC) The relationship between `Avg TiDB KV Request Duration` and `Avg TiKV GRPC Duration` is as follows: @@ -474,7 +474,7 @@ According to the preceding formula, the QPS of a write-heavy OLTP workload in v5 - v5.3.0: 24.4 ms ~= 17.7 ms + 6.59 ms - v5.4.0: 21.4 ms ~= 14.0 ms + 7.33 ms -In v5.4.0, the gPRC module has been optimized to accelerate Raft log replication, which reduces `Store Duration` compared with v5.3.0. +In v5.4.0, the gRPC module has been optimized to accelerate Raft log replication, which reduces `Store Duration` compared with v5.3.0. v5.3.0: @@ -521,7 +521,7 @@ The QPS of a write-heavy OLTP workload in v5.4.0 is improved by 14% compared wit | Commit Log Duration | 13 | 8.68 | | Apply Log Duration | 0.457|0.514 | -In v5.4.0, the gPRC module has been optimized to accelerate Raft log replication, which reduces `Commit Log Duration` and `Store Duration` compared with v5.3.0. +In v5.4.0, the gRPC module has been optimized to accelerate Raft log replication, which reduces `Commit Log Duration` and `Store Duration` compared with v5.3.0. v5.3.0: diff --git a/releases/release-2.0.6.md b/releases/release-2.0.6.md index 9dbb73443f984..0b94ee0ea60b4 100644 --- a/releases/release-2.0.6.md +++ b/releases/release-2.0.6.md @@ -23,7 +23,7 @@ On August 6, 2018, TiDB 2.0.6 is released. Compared with TiDB 2.0.5, this releas - Fix the issue that the `DROP USER` statement is incompatible with MySQL behavior in some cases [#7014](https://github.com/pingcap/tidb/pull/7014) - Fix the issue that statements like `INSERT`/`LOAD DATA` meet OOM after opening `tidb_batch_insert` [#7092](https://github.com/pingcap/tidb/pull/7092) - Fix the issue that the statistics fail to automatically update when the data of a table keeps updating [#7093](https://github.com/pingcap/tidb/pull/7093) - - Fix the issue that the firewall breaks inactive gPRC connections [#7099](https://github.com/pingcap/tidb/pull/7099) + - Fix the issue that the firewall breaks inactive gRPC connections [#7099](https://github.com/pingcap/tidb/pull/7099) - Fix the issue that prefix index returns a wrong result in some scenarios [#7126](https://github.com/pingcap/tidb/pull/7126) - Fix the panic issue caused by outdated statistics in some scenarios [#7155](https://github.com/pingcap/tidb/pull/7155) - Fix the issue that one piece of index data is missed after the `ADD INDEX` operation in some scenarios [#7156](https://github.com/pingcap/tidb/pull/7156) diff --git a/releases/release-4.0.16.md b/releases/release-4.0.16.md index 5787f7000fd5a..d301a3228b249 100644 --- a/releases/release-4.0.16.md +++ b/releases/release-4.0.16.md @@ -41,7 +41,7 @@ TiDB version: 4.0.16 + TiCDC - Add a tick frequency limit to EtcdWorker to prevent frequent etcd writes from affecting PD services [#3112](https://github.com/pingcap/tiflow/issues/3112) - - Optimize rate limiting control on TiKV reloads to reduce gPRC congestion during changefeed initialization [#3110](https://github.com/pingcap/tiflow/issues/3110) + - Optimize rate-limiting control on TiKV reloads to reduce gRPC congestion during changefeed initialization [#3110](https://github.com/pingcap/tiflow/issues/3110) ## Bug fixes diff --git a/releases/release-5.0.6.md b/releases/release-5.0.6.md index d457430bca6d1..9f7c1bf2d1bb6 100644 --- a/releases/release-5.0.6.md +++ b/releases/release-5.0.6.md @@ -42,7 +42,7 @@ TiDB version: 5.0.6 + TiCDC - - Optimize rate limiting control on TiKV reloads to reduce gPRC congestion during changefeed initialization [#3110](https://github.com/pingcap/ticdc/issues/3110) + - Optimize rate-limiting control on TiKV reloads to reduce gRPC congestion during changefeed initialization [#3110](https://github.com/pingcap/ticdc/issues/3110) - Add a tick frequency limit to EtcdWorker to prevent frequent etcd writes from affecting PD services [#3112](https://github.com/pingcap/ticdc/issues/3112) - Add the default configuration for `config.Metadata.Timeout` in Kafka sink [#3352](https://github.com/pingcap/tiflow/issues/3352) - Set the default value of `max-message-bytes` to `10M`, to reduce the probability that Kafka messages cannot be sent [#3081](https://github.com/pingcap/tiflow/issues/3081) diff --git a/releases/release-5.1.4.md b/releases/release-5.1.4.md index aea49c76ca4e9..62830677306d9 100644 --- a/releases/release-5.1.4.md +++ b/releases/release-5.1.4.md @@ -55,7 +55,7 @@ TiDB version: 5.1.4 - Add metrics for observing the remaining time of incremental scan [#2985](https://github.com/pingcap/tiflow/issues/2985) - Reduce the count of "EventFeed retry rate limited" logs [#4006](https://github.com/pingcap/tiflow/issues/4006) - Add more Prometheus and Grafana monitoring metrics and alerts, including `no owner alert`, `mounter row`, `table sink total row`, and `buffer sink total row` [#4054](https://github.com/pingcap/tiflow/issues/4054) [#1606](https://github.com/pingcap/tiflow/issues/1606) - - Optimize rate limiting control on TiKV reloads to reduce gPRC congestion during changefeed initialization [#3110](https://github.com/pingcap/ticdc/issues/3110) + - Optimize rate-limiting control on TiKV reloads to reduce gRPC congestion during changefeed initialization [#3110](https://github.com/pingcap/ticdc/issues/3110) - Reduce the time for the KV client to recover when a TiKV store is down [#3191](https://github.com/pingcap/tiflow/issues/3191) ## Bug fixes diff --git a/releases/release-6.1.0.md b/releases/release-6.1.0.md index fd03b15ae0d73..ac10b7f364376 100644 --- a/releases/release-6.1.0.md +++ b/releases/release-6.1.0.md @@ -342,7 +342,7 @@ In 6.1.0, the key new features or improvements are as follows: - CDC supports RawKV [#11965](https://github.com/tikv/tikv/issues/11965) - Support splitting a large snapshot file into multiple files [#11595](https://github.com/tikv/tikv/issues/11595) - Move the snapshot garbage collection from Raftstore to background thread to prevent snapshot GC from blocking Raftstore message loops [#11966](https://github.com/tikv/tikv/issues/11966) - - Support dynamic setting of the maximum message length (`max-grpc-send-msg-len`) and the maximum batch size of gPRC messages (`raft-msg-max-batch-size`) [#12334](https://github.com/tikv/tikv/issues/12334) + - Support dynamic setting of the maximum message length (`max-grpc-send-msg-len`) and the maximum batch size of gRPC messages (`raft-msg-max-batch-size`) [#12334](https://github.com/tikv/tikv/issues/12334) - Support executing online unsafe recovery plan through Raft [#10483](https://github.com/tikv/tikv/issues/10483) + PD diff --git a/releases/release-6.1.3.md b/releases/release-6.1.3.md index 2f39576e6123b..8abab08f3007b 100644 --- a/releases/release-6.1.3.md +++ b/releases/release-6.1.3.md @@ -44,7 +44,7 @@ Quick access: [Quick start](https://docs-archive.pingcap.com/tidb/v6.1/quick-sta - Fix the issue of the wrong query result that occurs when the mistakenly pushed-down conditions are discarded by Join Reorder [#38736](https://github.com/pingcap/tidb/issues/38736) @[winoros](https://github.com/winoros) - Fix the issue that the lock acquired by `get_lock()` cannot hold for more than 10 minutes [#38706](https://github.com/pingcap/tidb/issues/38706) @[tangenta](https://github.com/tangenta) - Fix the issue that the auto-increment column cannot be used with check constraint [#38894](https://github.com/pingcap/tidb/issues/38894) @[YangKeao](https://github.com/YangKeao) - - Fix the issue that the gPRC log is output to a wrong file [#38941](https://github.com/pingcap/tidb/issues/38941) @[xhebox](https://github.com/xhebox) + - Fix the issue that the gRPC log is output to a wrong file [#38941](https://github.com/pingcap/tidb/issues/38941) @[xhebox](https://github.com/xhebox) - Fix the issue that the TiFlash sync status of a table is not deleted from etcd when the table is truncated or dropped [#37168](https://github.com/pingcap/tidb/issues/37168) @[CalvinNeo](https://github.com/CalvinNeo) - Fix the issue that data files can be accessed unrestrainedly via data source name injection (CVE-2022-3023) [#38541](https://github.com/pingcap/tidb/issues/38541) @[lance6716](https://github.com/lance6716) - Fix the issue that the function `str_to_date` returns wrong result in the `NO_ZERO_DATE` SQL mode [#39146](https://github.com/pingcap/tidb/issues/39146) @[mengxin9014](https://github.com/mengxin9014)