… by simply not fsyncing every transaction (innodb_flush_log_at_trx_commit != 1), or work around it with by fsyncing to memory (Battery… no effect on standard MySQL replication from this instance, since MySQL replication is asynchronous. What about semi-sync MySQL replication? It’s actually…
Post: MySQL and Percona Server in LinkBench benchmark
May 8, 2013 By 18 Comments
…. Percona Server 5.5 outperforms MySQL in about 2x times. Both MySQL 5.5.30 and MySQL 5.6.11 demonstrate notable drops… we can see small performance drop in MySQL 5.6. In IO-bound cases MySQL still has performance issues around mutexes and… innodb_log_files_in_group = 2 innodb_flush_log_at_trx_commit = 1 innodb_log_buffer_size=128M innodb_max_dirty_pages_pct…
Post: More on MySQL transaction descriptors optimization
April 26, 2013 By Leave a Comment
…_data_file_path=ibdata1:100M:autoextend innodb_flush_log_at_trx_commit = 2 innodb_flush_method = O_DIRECT innodb_log_buffer_size…=0 –percentile=99 –mysql-user=root –mysql-db=sbtest8t1M –mysql-table-engine=INNODB –mysql-socket=/tmp/mysql.sock –oltp-point-selects=1 –oltp-simple-ranges=0…
Post: Benchmarking Percona Server TokuDB vs InnoDB
May 7, 2013 By 32 Comments
…/1000, $K) ON DUPLICATE KEY UPDATE k=VALUES(k)+k; COMMIT; So this workload produces SEQUENTIAL inserts into Primary Key, that…_DIRECT innodb_log_buffer_size = 256M innodb_flush_log_at_trx_commit = 1 innodb_buffer_pool_size = 36G innodb_log_file_size = 4G…_sort_file_size = 10G myisam_repair_threads = 1 myisam_recover socket=/var/lib/mysql/mysql.sock user=root skip-grant-tables TokuDB-related…
Post: Virident vCache vs. FlashCache: Part 2
May 17, 2013 By 3 Comments
…=1 –rand-type=pareto \ –forced-shutdown=1 –max-time=7200 –max-requests=0 –percentile=95 \ –mysql-user=root –mysql-socket=/tmp/mysql.sock –mysql-…log_at_trx_commit = 1 innodb_log_file_size = 1G innodb_log_files_in_group = 2 innodb_purge_threads = 1 innodb_fast_shutdown = 1 #…
Post: Memory allocators: MySQL performance improvements in Percona Server 5.5.30-30.2
April 15, 2013 By
… with trx_list scan we discussed in Friday’s post, there is another issue in InnoDB transaction processing that notably affects MySQL… each commit and thus the read view memory is reallocated on the next transaction. There are two aspects of this problem: 1… test for glibc and jemalloc allocators. Observations: = MySQL 5.5.30 – throughput of MySQL 5.5.30 with glibc is limited first…
Comment: trx descriptors: MySQL performance improvements in Percona Server 5.5.30-30.2
April 12, 2013 By 3 Comments
…1) The 1st benchmark is single SELECT queries (i.e. the –skip-trx option to sysbench). The 2nd one is the default sysbench mode, i.e. “BEGIN; SELECT …; COMMIT” which … There was no fix from Percona at the time MySQL bug #49169 was closed. However, as you can…
Post: Testing the Group Commit Fix
July 13, 2011 By 13 Comments
… recover-ability requirements, that is innodb_flush_log_at_trx_commit=1, sync_binlog=1 and you do not have storage that provides fast…), innodb_flush_log_at_trx_commit=1 and sync_binlog=1 or in tabular format: Threads Nofix Group Commit fix 1 21.51 21.99 2… improves performance substantially when both innodb_flush_log_at_trx_commit=1 and sync_binlog=1 options are set and write-back cache is…
Post: Reasons for MySQL Replication Lag
July 29, 2011 By 9 Comments
One common theme in the questions our MySQL Support customers ask is MySQL Replication Lag. The story is typically along the lines… for replication include setting sync_binlog=1, enabling log_slave_updates, setting innodb_flush_log_at_trx_commit=1. Also note there is no… as increasing buffer pool size could cause performance problems. MySQL Restart MySQL Restarts may include recovery time but what is the most…
Post: Maximal write througput in MySQL
February 28, 2010 By 10 Comments
…what maximal amount transactions per second we can get using MySQL and XtraDB / InnoDB storage engine if we have high-end…_at_trx_commit = 0 and binary logs trx_commit=1 & binlog : innodb_flush_log_at_trx_commit = 1 and binary logs trx_commit=1 & binlog & sync_bin : innodb_flush_log_at_trx_commit = 1…
