-
1
Dominik Grabiec - Techniques to Optimise Multi-threaded Data Building During Game Development - CppCon 2024Hello
My name is Dominik Grabiec
This talk is
Focusing on optimising the process around timestamp means recompute hash
▶ Use caches - for everything possible
Processing caches
Used internally during data building
Store frequently read values - save on IO & processing
Output caches
Store built data - referenced by raw pointer
Sorting needs to be quick - done many times
Both at construction and during queries
Speaker notesSORTING GRID NODES
inline bool operator<(const GridNode& left, const GridNode& 0 码力 |
99 页 |
2.40 MB
| 5 月前 3
-
· 1194 7.6.4 The start-ts timestamp of the TiCDC task is quite different from the current time. During the execution of this task, replication is in- terrupted and an error [CDC:ErrBufferReachLimit] occurs · · · · · 1195 7.6.8 How can I find out whether a DDL statement fails to execute in down- stream during TiCDC replication? How to resume the replication?· · · · 1196 7.6.9 The Kafka: client has run out log backup data to your backup storage, you can encrypt �→ the backup data to ensure its security during storage and �→ transmission. 2.2.1 Feature details 2.2.1.1 Performance • Introduce parallel
0 码力 |
6730 页 |
111.36 MB
| 9 月前 3
-
log backup data to your backup storage, you can encrypt �→ the backup data to ensure its security during storage and �→ transmission. 2.2.1 Feature details 2.2.1.1 Performance • Introduce parallel #8432 @MyonKeminta Before v8.4.0, when requesting TSO from PD, TiDB collects multiple TSO requests during a specific period and processes them in batches serially to decrease the number of Remote Procedure only be enabled or dis- abled. When it is enabled, TiDB applies a table-level scattering strategy during batch table creation. However, when creating hundreds of thousands of tables in a batch, this strategy
0 码力 |
6705 页 |
110.86 MB
| 9 月前 3
-
high NDV data #9196 @guo-shaoge Before v8.3.0, TiFlash has low aggregation calculation efficiency during the first stage of HashAgg aggregation when handling data with high NDV (number of distinct values) additional monitoring API so you can obtain the progress of loading initial statistics at any time during the startup phase. • Add metrics about Request Unit (RU) settings #8444 @nolouch 2.2.1.6 Security concur- rent threads during the Re- solve Locks and Delete Range steps of the Garbage Collec- tion (GC) pro- cess. Before v8.3.0, this vari- able only con- trols the num- ber of threads during the Re- solve Locks
0 码力 |
6606 页 |
109.48 MB
| 9 月前 3
-
recated. As an alternative, you can configure the maximum number of concur- rent tasks per TiKV node during snapshot restore using --tikv-max-restore- �→ concurrency. • Starting from v8.2.0, the BR snapshot table with a large amount of data (>1024 rows) #53871 @crazycs520 • Remove stores without Regions during MPP load balancing #52313 @xzhangxian1008 • TiKV 54 • Add the Compaction Job Size(files) metric and stability during node restarts, cluster scaling-out, and network jitter when backing up large numbers of tables #52534 @3pointer • Implement fine-grained checks of TiCDC changefeed during data restore
0 码力 |
6549 页 |
108.77 MB
| 9 月前 3
-
through text or execution plans within a period of time, to avoid excessive resource consumption during the identification phase when the concurrency of problematic queries is too high. TiDB also supports service automatically performs a forceRebase operation when the TiDB process exits (for example, during the TiDB node restart) to keep auto-assigned IDs as consecutive as possible. How- ever, when there the forceRebase behavior, but this change will cause some auto-assigned IDs to be non-consecutive during the failover. • In earlier versions, when processing a transaction containing UPDATE changes, if
0 码力 |
6479 页 |
108.61 MB
| 9 月前 3
-
peaks and valleys, you 39 can use this feature to perform data compaction during idle periods to improve the performance during peak periods. You can set the specific times that TiKV initiates periodic of the TiDB cluster more balanced and not affecting user access to the database during maintenance operations. • During maintenance operations such as rolling restarts, rolling upgrades, and scaling-in and the TiDB servers. By using TiProxy, connections can be smoothly migrated to other TiDB servers during these maintenance operations so that clients are not affected. • Client connections to a TiDB server
0 码力 |
6123 页 |
107.24 MB
| 1 年前 3
-