Multiprocess Queue

Queue in a threaded backend is changed to multiprocessing Queue to support sending metrics from processes in addition to threads

Async Decorator

Added

• Async decorator that can be used to log execution time of awaitable objects in Python >=3.5.

ElasticSearchBackend supporting ElasticSearch 5

CHANGED

• ElasticSearch index template is now created according to ElasticSearch 5 changes is not compatible with ElasticSearch 2

ElasticSearch backend fixes

Changed

• create ElasticSearch index following the index name pattern
• number of replicas in ElasticSearch template

Fixed

• lint error in threaded module

Non-daemonized threaded backend

time_execution.backends.threaded.ThreadedBackend is changed to start non-daemon thread for delivering metrics

Name the thread in threaded backend

Set the name to the thread in ThreadedBackend

Pin elasticsearch version to 2.1

Merge pull request #27 from sergray/master

Fix elasticsearch version

Fix worker TypeError

Merge pull request #26 from sergray/master

FIX workaround TypeError in queue module

FIX TypeError in threaded backend if queue is gone

Merge pull request #25 from sergray/master

FIX TypeError if the queue in ThreadedBackend is gone

Improve Elastic Search backend metrics indexing

Merge pull request #22 from sergray/master

FIX index origin metric attribute