June 19, 2013

Post: XtraDB benchmarks - 1.5X gain in IO-bound load

… in XtraDB are related to multi-core and multi-disks utilization. Also results may be different if load is CPU bound…_DIRECT innodb_file_per_table = 1 And for XtraDB I additionally used: innodb_io_capacity = 10000 innodb_adaptive_checkpoint = 1 innodb_write_io_threads = 16 innodb_read_io_threads…

Post: Effect from innodb log block size 4096 bytes

innodb_adaptive_checkpoint=keep_average innodb_thread_concurrency=0 innodb_flush_method = O_DIRECT innodb_read_ahead = none innodb_flush_neighbor_pages = 0 innodb_write_io_threads=16 innodb

Post: Innodb Fuzzy checkpointing woes

… value, ie 30% utilized at which we can stop checkpoint kind  flushes, which would avoid uneeded IO. One more related problem Innodb has – too many hard coded numbers.  For example Innodb

Post: InnoDB, InnoDB-plugin vs XtraDB on fast storage

… is better) : As you see InnoDB-plugin is doing much better here, it allows to utilize multiple IO threads, which as we saw… IO threads, we have patch to decrease contention on buffer_pool mutex, plus separate purge thread. Also we use different adaptive_checkpoint algorithm. The results are: So I guess buffer_pool improvements play here for XtraDB, and looking on summary result: InnoDB

Post: Linux schedulers in tpcc like benchmark

…vmstat) during benchmark As you see noop / deadline can utilize disks much better. For reference I used tpcc …innodb_log_file_size=128M innodb_thread_concurrency=0 innodb_flush_method = O_DIRECT innodb_write_io_threads=4 innodb_read_io_threads=4 innodb_io_capacity=800 innodb_adaptive_checkpoint

Post: XtraDB in CPU-bound benchmark

…they are very different for XtraDB and Innodb storage engines. Innodb has generally higher CPU usage and …dip in the middle. This happens because of checkpointing activity which can basically stall processing while … utilize CPU more fully XtraDB will show more gains, though not 1.5x as for IO bound…