- 14 Sep, 2018 2 commits
-
-
Daniele Venzano authored
-
Daniele Venzano authored
-
- 22 Jan, 2018 1 commit
-
-
Daniele Venzano authored
-
- 08 Jan, 2018 1 commit
-
-
Daniele Venzano authored
-
- 02 Jan, 2018 1 commit
-
-
Daniele Venzano authored
-
- 29 Nov, 2017 1 commit
-
-
Daniele Venzano authored
-
- 24 Nov, 2017 1 commit
-
-
Daniele Venzano authored
Small usability fixes for the execution inspect page Move all the database access in the state package Remove image management, it is too slow to download images before while the user waits the the ZApp to start Make several thread loops more robust in case of uncaught exceptions
-
- 21 Nov, 2017 2 commits
-
-
Daniele Venzano authored
-
Daniele Venzano authored
-
- 02 Nov, 2017 1 commit
-
-
- 23 Oct, 2017 1 commit
-
-
Daniele Venzano authored
-
- 20 Oct, 2017 2 commits
-
-
Daniele Venzano authored
-
Daniele Venzano authored
-
- 19 Oct, 2017 1 commit
-
-
Daniele Venzano authored
-
- 18 Oct, 2017 1 commit
-
-
Daniele Venzano authored
-
- 13 Oct, 2017 1 commit
-
-
Daniele Venzano authored
-
- 12 Oct, 2017 2 commits
-
-
Daniele Venzano authored
-
Daniele Venzano authored
-
- 11 Oct, 2017 3 commits
-
-
Daniele Venzano authored
-
Daniele Venzano authored
-
Daniele Venzano authored
This feature is implemented only on the DockerEnging back-end. When enabled via the new configuration option --backend-image-management, Zoe will (1) when an execution is submitted, start threads to pull the image on the nodes and (2) schedule services on the hosts that have the image already loaded. This has two big side-effects: 1. Zoe will not let Docker automatically pull the image at run time and will refuse to start a ZApp with unavailable images 2. ZApps will always run with the latest images: a pull is always done before a run
-
- 09 Oct, 2017 1 commit
-
-
Daniele Venzano authored
-
- 04 Oct, 2017 1 commit
-
-
Daniele Venzano authored
-
- 18 Sep, 2017 1 commit
-
-
Daniele Venzano authored
-
- 14 Sep, 2017 1 commit
-
-
Daniele Venzano authored
-
- 07 Sep, 2017 1 commit
-
-
Daniele Venzano authored
- remove executions with fatal errors form the scheduling queue - set services to error state when they fatally fail starting - re-implement log streaming that was not working
-
- 22 Mar, 2017 1 commit
-
-
Daniele Venzano authored
-
- 21 Mar, 2017 1 commit
-
-
Zoe Jenkins authored
Adds as a benefit: - support for TLS - no more monitor thread, as the event stream was unreliable
-
- 17 Mar, 2017 1 commit
-
-
hxquangnhat authored
-
- 13 Mar, 2017 2 commits
-
-
Quang-Nhat HOANG-XUAN authored
-
Quang-Nhat HOANG-XUAN authored
-
- 24 Feb, 2017 2 commits
-
-
Daniele Venzano authored
This is a temporary solution, we still need to find a way to handle service logs nicely
-
Daniele Venzano authored
-
- 31 Jan, 2017 3 commits
-
-
Daniele Venzano authored
-
Daniele Venzano authored
-
Daniele Venzano authored
-
- 30 Jan, 2017 1 commit
-
-
qhoangxuan authored
-
- 27 Jan, 2017 1 commit
-
-
Daniele Venzano authored
-
- 25 Jan, 2017 1 commit
-
-
qhoangxuan authored
-
- 24 Jan, 2017 1 commit
-
-
Daniele Venzano authored
-