Releases: criteo/blackbox-prober
Releases · criteo/blackbox-prober
blackbox-prober-0.1.31
blackbox-prober-0.1.30: Add configurable total timeout parameter
- Since we bumped from v5 to v7 of the Aerospike client the default timeout went from 0 (no timeout) to 1 second. - This change adds a configuration option to specify this value (total_timeout, e.g. "30s" in the yaml config). The default value is 30s - We also use explicitly (for readability) 2 retries for the durability checks (read and write policies): this is the default for read (aka. base) policy, while the default for write policiy is 0
blackbox-prober-0.1.29
What's Changed
Full Changelog: blackbox-prober-0.1.28...blackbox-prober-0.1.29
blackbox-prober-0.1.28
What's Changed
Full Changelog: blackbox-prober-0.1.27...blackbox-prober-0.1.28
blackbox-prober-0.1.27
What's Changed
- Remove all references to memcached by @damsallem in #41
- bump to aerospike client v7 by @geobeau in #42
Full Changelog: blackbox-prober-0.1.26...blackbox-prober-0.1.27
blackbox-prober-0.1.26
Clean the old 'aerospike-namespaces' consul metadata that is no longer supported
blackbox-prober-0.1.25
this release fix the issue regarding the 0.1.24 that failed to parse the new MetaKey added (aerospike-monitoring-)
blackbox-prober-0.1.24: Support new MetaData aerospike-monitoring-*
Following a consul limitation reached when using the metadata "aerospike-namespaces" (max for metadata value is 512 bytes), we needed to split into multiple metadata (1 per namespace) to activate the monitoring.
blackbox-prober-0.1.23: Allow configuring ExitFastOnExhaustedConnectionPool
The default behavior of the aerospike client is to loop forever trying to openconnections when a node is down. Now, if it tried to open too many connections it will stop right aways
blackbox-prober-0.1.22: Remove warmup of the aerospike client
It spawns a very high number of go routine. Instead minConnection is prefered.