Documentation Home
MySQL 5.7 Reference Manual
Related Documentation Download this Manual
PDF (US Ltr) - 35.3Mb
PDF (A4) - 35.5Mb
PDF (RPM) - 33.8Mb
EPUB - 8.6Mb
HTML Download (TGZ) - 8.4Mb
HTML Download (Zip) - 8.5Mb
HTML Download (RPM) - 7.3Mb
Eclipse Doc Plugin (TGZ) - 9.3Mb
Eclipse Doc Plugin (Zip) - 11.3Mb
Man Pages (TGZ) - 197.9Kb
Man Pages (Zip) - 301.8Kb
Info (Gzip) - 3.2Mb
Info (Zip) - 3.2Mb
Excerpts from this Manual Tuning InnoDB memcached Plugin Performance

Because using InnoDB in combination with memcached involves writing all data to disk, whether immediately or sometime later, raw performance is expected to be somewhat slower than using memcached by itself. When using the InnoDB memcached plugin, focus tuning goals for memcached operations on achieving better performance than equivalent SQL operations.

Benchmarks suggest that queries and DML operations (inserts, updates, and deletes) that use the memcached interface are faster than traditional SQL. DML operations typically see a larger improvements. Therefore, consider adapting write-intensive applications to use the memcached interface first. Also consider prioritizing adaptation of write-intensive applications that use fast, lightweight mechanisms that lack reliability.

Adapting SQL Queries

The types of queries that are most suited to simple GET requests are those with a single clause or a set of AND conditions in the WHERE clause:

SELECT col FROM tbl WHERE key = 'key_value';

GET key_value

SELECT col FROM tbl WHERE col1 = val1 and col2 = val2 and col3 = val3;

# Since you must always know these 3 values to look up the key,
# combine them into a unique string and use that as the key
# for all ADD, SET, and GET operations.
key_value = val1 + ":" + val2 + ":" + val3
GET key_value

SELECT 'key exists!' FROM tbl
  WHERE EXISTS (SELECT col1 FROM tbl WHERE KEY = 'key_value') LIMIT 1;

# Test for existence of key by asking for its value and checking if the call succeeds,
# ignoring the value itself. For existence checking, you typically only store a very
# short value such as "1".
GET key_value

Using System Memory

For best performance, deploy the daemon_memcached plugin on machines that are configured as typical database servers, where the majority of system RAM is devoted to the InnoDB buffer pool, through the innodb_buffer_pool_size configuration option. For systems with multi-gigabyte buffer pools, consider raising the value of innodb_buffer_pool_instances for maximum throughput when most operations involve data that is already cached in memory.

Reducing Redundant I/O

InnoDB has a number of settings that let you choose the balance between high reliability, in case of a crash, and the amount of I/O overhead during high write workloads. For example, consider setting the innodb_doublewrite to 0 and innodb_flush_log_at_trx_commit to 2. Measure performance with different innodb_flush_method settings.


innodb_support_xa is deprecated and will be removed in a future release. As of MySQL 5.7.10, InnoDB support for two-phase commit in XA transactions is always enabled and disabling innodb_support_xa is no longer permitted.

For other ways to reduce or tune I/O for table operations, see Section 9.5.8, “Optimizing InnoDB Disk I/O”.

Reducing Transactional Overhead

A default value of 1 for daemon_memcached_r_batch_size and daemon_memcached_w_batch_size is intended for maximum reliability of results and safety of stored or updated data.

Depending on the type of application, you might increase one or both of these settings to reduce the overhead of frequent commit operations. On a busy system, you might increase daemon_memcached_r_batch_size, knowing that changes to data made through SQL may not become visible to memcached immediately (that is, until N more get operations are processed). When processing data where every write operation must be reliably stored, leave daemon_memcached_w_batch_size set to 1. Increase the setting when processing large numbers of updates intended only for statistical analysis, where losing the last N updates in a crash is an acceptable risk.

For example, imagine a system that monitors traffic crossing a busy bridge, recording data for approximately 100,000 vehicles each day. If the application counts different types of vehicles to analyze traffic patterns, changing daemon_memcached_w_batch_size from 1 to 100 reduces I/O overhead for commit operations by 99%. In case of an outage, a maximum of 100 records are lost, which may be an acceptable margin of error. If instead the application performed automated toll collection for each car, you would set daemon_memcached_w_batch_size to 1 to ensure that each toll record is immediately saved to disk.

Because of the way InnoDB organizes memcached key values on disk, if you have a large number of keys to create, it may be faster to sort the data items by key value in the application and add them in sorted order, rather than create keys in arbitrary order.

The memslap command, which is part of the regular memcached distribution but not included with the daemon_memcached plugin, can be useful for benchmarking different configurations. It can also be used to generate sample key/value pairs to use in your own benchmarks. See Section, “libmemcached Command-Line Utilities” for details.

User Comments
Sign Up Login You must be logged in to post a comment.