|
Revision tags: v7.0.2, v6.6.3, v7.0.1, v7.0.0, v6.6.2, v6.5.2, v6.6.1, v6.0.5, v6.6.0, v6.5.1, v6.0.4, v6.5.0 |
|
| #
5e3f2c1d |
| 27-Nov-2019 |
Trond Norbye <trond.norbye@gmail.com> |
MB-36923: Allow "default" and "disk_io_optimized" for ThreadCount
Allow the values "default" and "disk_io_optimized" to be used for the
num_reader_threads & num_writer_threads memcached.
MB-36923: Allow "default" and "disk_io_optimized" for ThreadCount
Allow the values "default" and "disk_io_optimized" to be used for the
num_reader_threads & num_writer_threads memcached.json settings:
* "default": Set the number of reader/writer threads to be a balanced
value which is reasonable for most workloads:
- num_writers: 4
- num_readers: Number of available CPU cores; minimum 4, maximum 16.
This is broadly similar to the 6.0 default calculated values.
* "disk_io_optimized": Set the number of reader/writer threads to be
optimised for Disk IO latency:
- num_writers: Number of available CPU cores; minimum 4, maximum 64.
- num_readers: Number of available CPU cores; minimum 4, maximum 64.
Change-Id: I56c62e624f7f067168a2b074fb079602d1287983
Reviewed-on: http://review.couchbase.org/118511
Well-Formed: Build Bot <build@couchbase.com>
Reviewed-by: Jim Walker <jim@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>
show more ...
|
|
Revision tags: v6.0.3, v5.5.4, v5.5.5, v5.5.6 |
|
| #
2224a71f |
| 06-Feb-2019 |
Paolo Cocchi <paolo.cocchi@couchbase.com> |
MB-32904 [SR]: Introduce the DurabilityTimeoutTask
This patch introduces the DurabilityTimeoutTask, which is responsible
for enforcing the Durability Timeout for the SyncWrites tracked i
MB-32904 [SR]: Introduce the DurabilityTimeoutTask
This patch introduces the DurabilityTimeoutTask, which is responsible
for enforcing the Durability Timeout for the SyncWrites tracked in
KVBucket.
With this patch the Timeout is enforced internally in KVEngine, i.e.:
- the task runs periodically and checks for timed-out SyncWrites
- the timed-out SyncWrites are removed from tracking, so they will
never be committed
But, given that VBucket::abort is not implemented yet, then:
- timed-out SyncWrites will stale in HashTable/disk as pending
- clients will hang as we do not send any DCP_ABORT yet
Change-Id: I71db84426f829f8acd4d83a82b73f0d381fb1ed6
Reviewed-on: http://review.couchbase.org/104638
Reviewed-by: Dave Rigby <daver@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>
show more ...
|