… of the default 128M Rationale: Since O_DIRECT is not used by default, the file system cache …12GB memory 4 disk 7200RPM RAID 10 array with 512MB write-back cache Star Schema Benchmark – Scale Factor…_EVICTED_RATE: 0 LRU_IO_TOTAL: 531 * LRU_IO_TOTAL: 431 LRU_IO_CURRENT: 0 * LRU_IO_CURRENT: 0 …
Post: SSD, XFS, LVM, fsync, write cache, barrier and lost transactions
March 2, 2009 By 32 Comments
…commit=1 innodb_flush_method = O_DIRECT Actually most interesting one are …all committed transactions even in case of system failure. Running this test …’s SSD, we should have much more IO there. Investigations put me into … put SSD into RAID controller with battery-backup-ed write cache, but I am…
Post: How much overhead DRDB could cause ?
June 2, 2008 By 4 Comments
… should expect. First lets talk what kind of IO you performance you care about while running…RAID as I benchmarked you can be getting over 10000 req/sec in-cache write speed (and this is what a lot of… rather interesting to see DRBD to have direct support for Infiniband or Dolphin interconnect sockets…
Post: Should MySQL and Web Server share the same box ?
October 16, 2006 By 9 Comments
… Compromising your web server does not directly give access to your database…to avoid database corruption, use RAID to avoid loosing database …find database servers to be IO bound so CPUs are …caching – Web page caching, if you do not have separate layer for it, local memory or cacheing type of caching…
Comment: Evaluating IO subsystem performance for MySQL Needs
March 5, 2008 By 23 Comments
…cache they can be lost on power down anyway. If you’re speaking about unbuffered IO to device (O_SYNC, O_DIRECT… incorrectly (ie forced Writeback cache mode on the RAID without BBU) and …cache is removed from equation because of use of O_DIRECT which bypasses OS cache for reads and writes. 256MB cache…
Post: Innodb Performance Optimization Basics
November 1, 2007 By 75 Comments
… lot. The third is IO Subsystem – directly attached storage with plenty of spindles and RAID with battery backed up cache is a good bet. Typically… careful if you do not have battery backed up RAID cache as when write IO may suffer. innodb_file_per_table – If you… be better of when majority of Innodb users which take MySQL with defaults run it on hardware without battery backed up cache with…
Post: FlashCache: tpcc workload
May 25, 2010 By 18 Comments
…start. Short version of results are on graph: In summary: on RAID final result:…does not work (you need O_DIRECT with proper blocksize), and you … FlashCache was able to attach previous cache, and MySQL was able to … reads this post – please change default IO scheduler from CFQ to Deadline. Seriously…
Comment: Evaluating IO subsystem performance for MySQL Needs
March 5, 2008 By 23 Comments
… when using O_DIRECT-deadline should win. What is your IO scheduler setting? Most real-world cases the Cache on the controller is… skewed by the system cache? So, why choose RAID-5 for this test, when RAID-10 is perfect for random IO? For testing disks… throughput to verify your test is testing the RAID-setup itself and not the various OS Layers in front of the disks?
Post: Benchmarking single-row insert performance on Amazon EC2
May 16, 2012 By 17 Comments
…for the IO configuration I chose 8 x 200G EBS volumes in software RAID 10. …DIRECT innodb_max_dirty_pages_pct = 50 innodb_io_capacity = 800 innodb_read_io_threads = 8 innodb_write_io…cache to avoid contention caused by query cache on write heavy workload. OK, so that was all about the configuration of…
Comment: How to decrease InnoDB shutdown times
April 15, 2009 By 20 Comments
…InnoDB to lower the percentage of dirty buffers, it is …io_capacity, more background IO threads and a few others. Writes to files in the OS buffer cache are fast. Writes may be slow when: * using innodb_flush_method=O_DIRECT * disabling SATA writeback cache * not using a HW RAID disk cache…
