May 25, 2012

Post: Benchmarking single-row insert performance on Amazon EC2

…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_iocache to avoid contention caused by query cache on write heavy workload. OK, so that was all about the configuration of

Post: Aligning IO on a hard disk RAID – the Benchmarks

… during “prepare” phase and instead of having files created using direct IO they were buffered in OS cache and so writes to files were… on. As I mentioned earlier, each of those were run twice – first with RAID controller cache set to Write-Through and then to… regardless of file size, let’s look at them. First, results with WT cache: And here’s with WB cache: Apparently proper IO alignment…

Post: Query Profiling with MySQL: Bypassing caches

…_DIRECT flag or mounting file system in direct IO mode. Hardware Caches State of CPU cache may affect query execution speed but only lightly, the hardware IO cache may however cause dramatic difference. Hardware RAID cache

Post: FlashCache: first experiments

… rows (~18GB of data) And I am comparing results with data stored: Directly on RAID Directly on SSD Data stored on RAID, but caching in FlashCache… I should mention that FlashCache uses WriteBack caching algorithm, which keeps some amount of dirty pages before physically writing to main storage… see how it affects ( expected that smaller amount of dirty pages puts more IO on main storage leading to smaller performance). So…

Post: The performance effects of new patches

… is enough bigger than the data cache of the storage. main common settings innodb…DIRECT The next graphs show the frequency of IO, and you can see we can get more IOio-threads.patch Generally RAID storage can parallelize IO accesses to several disks. This patch enables to control the number of the IO

Post: Heikki Tuuri Innodb answers - Part I

… here. Memlock and O_DIRECT. Memlock is apparently unstable on Linux. O_DIRECT is an alternative but Linus… to reduce number of TLB cache misses. huge pages along however do not stop IO pressure and OS may….0 because of group commit being broken. Most do not have RAID with BBU. With good RAID with BBU…

Post: Choosing innodb_buffer_pool_size

…bypass OS buffering Innodb TableSpace IO will wipe out cache because it typically makes most of the IO on the system. … Solaris you need to set innodb_flush_method=O_DIRECT. On other Operating Systems you may be able … when it hurts – when you do not have RAID with BBU and your workload is very write …

Post: Evaluating IO subsystem performance for MySQL Needs

of memory (or reducing amount of usable memory). The system I’m testing this on has 256MB BBU (Battery Backed up Cache) on RAID… both by RAID cache and uncached IO performance. We’re interested in cached reads because they may show latency to the RAID cache if RAID read cache is… where. The problem is basically in case of O_DIRECT writes there is only one IO can be happening per file at any…

Post: Dissection of EC2 / EBS volume

RAID as in my previous post). Yasufumi did some benchmarks and pointed me on interesting behavior, there seems several level of caching on EBS volume. Let me show you. I did sysbench random read IO …=on \ –file-num=1 –file-extra-flags=direct –file-fsync-freq=0 run \ >> sysbench-…

Post: When would you use SAN with MySQL ?

…some very smart software instead of simple RAID you use with directly attach storage which can magically…caching. SAN can have a lot of cache though servers can typically have more. If you can afford SAN you should afford 128GB of…slave thread will issue single IO request most of the time making it …