faktory_worker_ex changes
0.7.1
0.7.0
- The great simplication refactor; everything done in this version is to reduce conceptual complexity and make the code easier to read and understand.
- Changed to a queue based architecture (reduces number of connections to Faktory server).
- No more Genservers, the components of a worker (producer, consumers, reporter) are all supervised tasks. Jobs are also run in tasks (spawned by the consumers).
- Removed
retryable_ex
; workers will retry when fetching, acking, and failing jobs, but all else is up to the user now.
0.6.0
- (breaking) Configuration simplified (again); see readme.
- Can override
jobtype
for enqueuing to workers in other languages.
- Simplified supervision; removed all custom supervisor modules.
- Workers no longer use connection pool; each processor has dedicated (linked) connection.
- Simplify plain vs tls socket connections with the
socket
package.
- Protocol now retries with the
retryable_ex
package.
- Faktory 0.9.3
0.5.0
- (breaking) Simplified middleware; no longer pass the
chain
arg.
- (breaking) Completely rewrote configuration system. See
Faktory.Configuration
.
- (breaking) Removed
set/1
and set/2
in favor of passing options
to perform_async/2
.
- (breaking)
perform_async/2
argument order reversed (options
is 2nd arg now).
- Can enqueue to different Faktory servers via
perform_async/2
.
0.4.0
- Faktory 0.6.x support @acj @valo
- Authentication support @acj
- TLS support @acj
- The
:retries
option was renamed to :retry
@valo
- Handle non-exception based errors (
{:EXIT, pid, :killed}
) @valo
- More tests and test specific "callbacks" to aid in testing.
- Update protocol to handle
$-1\r\n
@valo
0.3.0
- Configure with Mix Config and callback.
- Independently configurable logger.
- Connection pooling for workers.
0.2.1
- Start having a changelog.