Use bisection to find the position within the ring, as in ketama.
According to my benchmarking this is faster than the guess/refine
algorithm for rings with 100 locations, i.e. 16000 ring nodes requiring
at most 14 bisection iterations.
Change-Id: I6042f382de093081971b5ec210eda8dfa74988f2
This scheme avoids disruptive re-hashing when a node is
ejected; only the adjectent clockwise sector is affected,
the later taking responsibility for the former. Addition
of new nodes is likewise less disruptive.
When used with MD5, this maps keys the same as libketama.
Also:
* Make HashRing::getLocations() faster by guessing the
sector list index and moving clockwise/counter-clockwise.
* Made HashRing Serializable for more robust storage.
* Allow for different hash functions to be specified.
Note:
* This makes PoolCounterRedis use the new hash scheme.
* Likewise for JobQueueFederated (for job de-duplication).
Switching from one mode to the other will cause some of
duplicated effort from old jobs only temporarily.
Change-Id: Iff432a4e5dc3fb40b2dda70fb3dbe9dda0b74933
It's deprecated in PHP 7.2, may as well replace it now.
I note that, contrary to claims at
https://wiki.php.net/rfc/deprecations_php_7_2#each, none of our uses
were trivially replaceable with foreach.
* wfArrayDiff2_cmp() is processing two arrays by value in parallel.
* MagicWordArray::parseMatch() is doing something funky with the data
structure returned by preg_match().
* HashRing was using it like "nextKey()", replaced with calls to key()
and next().
* FormatMetadata and IndexPager were both using it as a shorter way to
get both key() and current() for the first element in the array. I
suppose a foreach(){ break; } would do the same, but that's confusing.
Bug: T174354
Change-Id: I36169a04c764fdf1bfd6603395111c6fe0aae5eb
The used phpcs has a bug, so the version 0.9.0 could not be enforced at the moment.
Will be fixed in next version, see T167168
Changed:
- Remove duplicate newline at end of file
- Add space between function and ( for closures
- and -> &&, or -> ||
Change-Id: I4172fb08861729bccd55aecbd07e029e2638d311
Most of these are simply changing annotations to reflect
reality. If a function can return false to indicate failure
the @return should indicate it.
Some are fixing preg_match calls, preg match returns 1, 0 or false,
but the functions all claim to return booleans.
This is far from all the incorrect return types in mediawiki, there
are around 250 detected by phan, but have to start somewhere.
Change-Id: I1bbdfee6190747bde460f8a7084212ccafe169ef
* Added proper ejection and caching in HashRing to avoid rebuilding
the hash all the time (or doing manual caching).
* Made JobQueueFederated blacklist failing servers for a few seconds.
* Also made the JobQueueFederated root job methods properly respect
the weights when they fail over.
Change-Id: Ifa4c03272c1777cfff2523ab21f780074ddcf359
