[ML] Don't install empty ML metadata on startup#30751
Merged
droberts195 merged 1 commit intoelastic:masterfrom May 21, 2018
Merged
[ML] Don't install empty ML metadata on startup#30751droberts195 merged 1 commit intoelastic:masterfrom
droberts195 merged 1 commit intoelastic:masterfrom
Conversation
This change is to support rolling upgrade from a pre-6.3 default distribution (i.e. without X-Pack) to a 6.3+ default distribution (i.e. with X-Pack). The ML metadata is no longer eagerly added to the cluster state as soon as the master node has X-Pack available. Instead, it is added when the first ML job is created. As a result all methods that get the ML metadata need to be able to handle the situation where there is no ML metadata in the current cluster state. They do this by behaving as though an empty ML metadata was present. This logic is encapsulated by always asking for the current ML metadata using a static method on the MlMetadata class. Relates elastic#30731
Collaborator
|
Pinging @elastic/ml-core |
droberts195
commented
May 21, 2018
| ClusterState state = clusterService.state(); | ||
| MlMetadata mlMetadata = state.getMetaData().custom(MLMetadataField.TYPE); | ||
|
|
||
| // Handle case when usage is called but MlMetadata has not been installed yet |
Author
There was a problem hiding this comment.
It was never really worth handling this case differently to ML metadata existing but being empty, as this state only existed for a fraction of a second on first cluster startup. Now this state will exist for longer - until the first job is created - but handling it like empty ML metadata is actually closer to the historical behaviour.
dimitris-athanasiou
approved these changes
May 21, 2018
Contributor
dimitris-athanasiou
left a comment
There was a problem hiding this comment.
LGTM This is so much cleaner!
droberts195
added a commit
that referenced
this pull request
May 21, 2018
This change is to support rolling upgrade from a pre-6.3 default distribution (i.e. without X-Pack) to a 6.3+ default distribution (i.e. with X-Pack). The ML metadata is no longer eagerly added to the cluster state as soon as the master node has X-Pack available. Instead, it is added when the first ML job is created. As a result all methods that get the ML metadata need to be able to handle the situation where there is no ML metadata in the current cluster state. They do this by behaving as though an empty ML metadata was present. This logic is encapsulated by always asking for the current ML metadata using a static method on the MlMetadata class. Relates #30731
droberts195
added a commit
that referenced
this pull request
May 21, 2018
This change is to support rolling upgrade from a pre-6.3 default distribution (i.e. without X-Pack) to a 6.3+ default distribution (i.e. with X-Pack). The ML metadata is no longer eagerly added to the cluster state as soon as the master node has X-Pack available. Instead, it is added when the first ML job is created. As a result all methods that get the ML metadata need to be able to handle the situation where there is no ML metadata in the current cluster state. They do this by behaving as though an empty ML metadata was present. This logic is encapsulated by always asking for the current ML metadata using a static method on the MlMetadata class. Relates #30731
9 tasks
jasontedor
added a commit
to jasontedor/elasticsearch
that referenced
this pull request
May 21, 2018
…ne-liners * elastic/master: [TEST] Enable DEBUG logging on testAutoQueueSizingWithMax [ML] Don't install empty ML metadata on startup (elastic#30751) Add assertion on removing copy_settings (elastic#30748) bump lucene version for 6_3_0 [DOCS] Mark painless execute api as experimental (elastic#30710) disable annotation processor for docs (elastic#30610) Add more script contexts (elastic#30721) Fix default shards count in create index docs (elastic#30747)
dnhatn
added a commit
that referenced
this pull request
May 21, 2018
* master: Reduce CLI scripts to one-liners (#30759) SQL: Preserve scoring in bool queries (#30730) QA: Switch rolling upgrade to 3 nodes (#30728) [TEST] Enable DEBUG logging on testAutoQueueSizingWithMax [ML] Don't install empty ML metadata on startup (#30751) Add assertion on removing copy_settings (#30748) bump lucene version for 6_3_0 [DOCS] Mark painless execute api as experimental (#30710) disable annotation processor for docs (#30610) Add more script contexts (#30721) Fix default shards count in create index docs (#30747) Mute testCorruptFileThenSnapshotAndRestore
martijnvg
added a commit
to martijnvg/elasticsearch
that referenced
this pull request
May 22, 2018
* es/ccr: (50 commits) Reduce CLI scripts to one-liners (elastic#30759) SQL: Preserve scoring in bool queries (elastic#30730) QA: Switch rolling upgrade to 3 nodes (elastic#30728) [TEST] Enable DEBUG logging on testAutoQueueSizingWithMax [ML] Don't install empty ML metadata on startup (elastic#30751) Add assertion on removing copy_settings (elastic#30748) bump lucene version for 6_3_0 [DOCS] Mark painless execute api as experimental (elastic#30710) disable annotation processor for docs (elastic#30610) Add more script contexts (elastic#30721) Fix default shards count in create index docs (elastic#30747) Mute testCorruptFileThenSnapshotAndRestore Scripting: Remove getDate methods from ScriptDocValues (elastic#30690) Upgrade to Lucene-7.4.0-snapshot-59f2b7aec2 (elastic#30726) [Docs] Fix single page :docs:check invocation (elastic#30725) Docs: Add uptasticsearch to list of clients (elastic#30738) [DOCS] Removes out-dated x-pack/docs/en/index.asciidoc [DOCS] Removes redundant index.asciidoc files (elastic#30707) [TEST] Reduce forecast overflow to disk test memory limit (elastic#30727) Plugins: Remove meta plugins (elastic#30670) ...
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This change is to support rolling upgrade from a pre-6.3 default
distribution (i.e. without X-Pack) to a 6.3+ default distribution
(i.e. with X-Pack).
The ML metadata is no longer eagerly added to the cluster state
as soon as the master node has X-Pack available. Instead, it
is added when the first ML job is created.
As a result all methods that get the ML metadata need to be able
to handle the situation where there is no ML metadata in the
current cluster state. They do this by behaving as though an
empty ML metadata was present. This logic is encapsulated by
always asking for the current ML metadata using a static method
on the MlMetadata class.
Relates #30731