Currently the runtime uses a compatibility mode for its reload strategy. This mode is unstash:
org.nuxeo.runtime.reload_strategy=unstash
Depending on runtime state, this mode will:
- stop components to remove
- deactivate components to remove
- apply the stash (components to remove/add before restarting)
- activate new components (<-- check if it's needed)
- activate new components and put them to desired state list (standby or started)
- start new components
There's two other modes standby and restart.
standby will:
- stop all components
- apply the stash
- start all components
restart will:
- stop all components
- deactivate all components
- apply the stash
- activate all components
- start all components
We should use standby mode, remove unstash mode and keep restart.
- depends on
-
NXBT-1803 Add a way to append template to existing list
- Resolved
-
NXP-21987 Avoid keeping reference to out-dated services
- Resolved
-
NXP-20847 Upgrade nuxeo-java-client version to 3.0.0
- Resolved
- is related to
-
NXP-30730 Fix potential NPE in ContributionFragmentRegistry#toMap
- Resolved
-
NXCLI-34 Add HotReload Integration Test on CLI
- Acknowledged
-
NXDOC-1520 Fix documentation about hot reload strategy
- Resolved
-
NXP-19326 Nuxeo Runtime Evolutions - Step 1
- Resolved
-
NXP-22860 Working hot reload strategy should not require a nuxeo.conf parameter
- Resolved
-
NXP-23145 Fix DevBundleTest.testParseDevBundleLines on windows
- Resolved
-
NXP-22905 Upgrade to org.nuxeo:nuxeo-ftest:1.8.3
- Resolved
- is required by
-
NXP-22937 Optimize Hot Reload
- Resolved
-
NXP-9642 Fix hot reload of IDE dev bundles resources
- Resolved
-
NXP-22704 Fix random UT failure on nuxeo-core WorkManagerTest.transientFilesWorkAreCleaned
- Resolved
-
NXP-22930 Add more hot reload integration tests
- Resolved
-
NXP-22986 Remove reload_strategy parameter from docker image
- Resolved