June 19, 2013

Post: The write cache: Swap insanity tome III

Swapping has always been something bad for MySQL performance but it is even more important for HA systems. It … (~150MB/s) but doing very little physical writes. The writes were cached and the write cache was causing the problem. It happened that… to 20 on a 48GB server, allowing nearly 10GB of “write cache”. A quick survey of the servers I have access…

Post: Why MySQL Performance at Low Concurrency is Important

… an opposite question: why MySQL performance at low concurrency is important for you. I decided to write about this topic as a number of recent blog posts and articles look at MySQL performance starting with certain… look as though single-thread performance is no more a concern. I believe it still is! First, performance at concurrency of 1…

Post: Is Synchronous Replication right for your app?

… the original node where the transaction is committed) Galera prevents writing conflicts to these pending transactions while they are inflight in… does NOT set an lower bound on performance, only a upper bound.  The best performance you can expect is modifying a given… update much faster.  I see this tuning very frequently for “performance” reasons when data durability isn’t as crucial.  This is…

Post: Virident vCache vs. FlashCache: Part 2

…, the baseline configuration plus: innodb_read_io_threads = 16 innodb_write_io_threads = 16 We call this configuration B. The next… – SSD config) which shows the most consistent performance. The others all have assorted performance drops scattered throughout the graph. We see the… time-based flushing enabled actually produces better and more consistent performance than vCache without time-based flushing, but even at its…

Post: Galera Flow Control in Percona XtraDB Cluster for MySQL

…. It contains no other magic that helps single node writing in PXC to perform better. gcs.fc_factor If fc_limit controls… large receive queue might be the result of some brief performance issue on a given node, or perhaps the effect of… nodes will be the slaves with no writes coming from them. However, with multi-node writing, larger queues will make certification more…

Comment: Virident vCache vs. FlashCache: Part 1

… only 500GB. What are my options for achieving the best performance? I could by a ton of RAM, but that isn…, not Virident, but there’s a comparison in TPCC-MySQL performance between that card and a traditional SSD – the PCIe card… speed than a regular SSD or HDD RAID), I can write data out to spinning disk at a nice sustainable rate…

Post: Benchmarking Percona Server TokuDB vs InnoDB

… Server with TokuDB, of course I wanted to compare InnoDB performance vs TokuDB. I have a particular workload I’m interested… how both InnoDB and TokuDB performs if PK is (`id`,`hid`,`mid`). This also will affect select performance, so we will need…_size=4096 #####plugin options innodb_read_io_threads = 16 innodb_write_io_threads = 4 innodb_io_capacity = 4000 innodb_use_native…

Post: MySQL and Percona Server in LinkBench benchmark

… MySQL performs quite well, though we can see small performance drop in MySQL 5.6. In IO-bound cases MySQL still has performance… innodb_file_per_table innodb_read_io_threads = 8 innodb_write_io_threads = 8 innodb_io_capacity = 5000 sync_binlog=0…-cache=1000 query_cache_size=0 query_cache_type=0 performance_schema=0 #56only loose-innodb_flush_neighbors=0 loose-metadata…

Comment: Is Synchronous Replication right for your app?

Excellent description. I have been waiting for someone to begin writing about Galera at this level of detail. I have not… semi-sync now so we will verify this. “Semi-sync performs dismally under concurrency. Think about it, our single inserts took… replication. Apparently each client must wait for it’s turn writing to the binlog and waiting for confirmation from the slave…

Post: Follow these basics when migrating to Percona XtraDB Cluster for MySQL

… consistent within the cluster. Also, at the time of this writing, from async stream, MyISAM is not being replicated at all… to the new cluster especially if you are planning to write on multiple nodes. Have PRIMARY KEYS If you still have…), table without PK imposes an additional overhead, and because cluster performance is somewhat dependent on the slowest node – this overhead can…