aboutsummaryrefslogtreecommitdiffstats
path: root/node-repository/src/main/java/com/yahoo/vespa/hosted/provision/autoscale
Commit message (Collapse)AuthorAgeFilesLines
* Remove unused argumentHarald Musum2024-03-152-8/+6
|
* Handle any exception typeJon Bratseth2024-01-301-1/+1
|
* Added comment, TODO, and better namingOla Aunronning2024-01-241-2/+2
|
* Store multiple resource suggestionsOla Aunronning2024-01-122-22/+59
|
* Handle missing value in new columnsMartin Polden2023-12-201-2/+7
|
* Store GPU metrics and load separatelyMartin Polden2023-12-205-49/+87
|
* Revert "Expose GPU load in nodes API"Harald Musum2023-12-195-87/+49
|
* Store GPU metrics and load separatelyMartin Polden2023-12-195-49/+87
|
* Remove unused parametersMartin Polden2023-12-081-3/+3
|
* Add TODOMartin Polden2023-12-071-0/+1
|
* Remove unused methodsMartin Polden2023-12-062-15/+0
|
* Avoid null optionalMartin Polden2023-12-062-3/+3
|
* Let distributor metric decide cluster stabilityMartin Polden2023-12-061-2/+20
|
* Revert "Evaluate flags referenced multiple times at the start of prepare"Håkon Hallingstad2023-11-075-60/+57
|
* Review fixesHåkon Hallingstad2023-11-062-3/+3
|
* Include NodeRepository, ApplcationId, and ClusterSpec as parametersHåkon Hallingstad2023-11-055-46/+29
|
* Rename to AllocationParamsHåkon Hallingstad2023-11-055-15/+15
|
* Rename to ClusterAllocationParamsHåkon Hallingstad2023-11-045-36/+36
|
* Evaluate flags referenced multiple times at the start of prepareHåkon Hallingstad2023-11-045-27/+47
|
* Update ↵Bjørn Meland2023-11-031-2/+2
| | | | | node-repository/src/main/java/com/yahoo/vespa/hosted/provision/autoscale/Autoscaler.java Co-authored-by: Jon Bratseth <bratseth@gmail.com>
* Clarification of messagebjormel2023-11-021-1/+1
|
* Don't autoscale single node clustersbjormel2023-11-021-1/+3
|
* Update copyrightJon Bratseth2023-10-0919-18/+19
|
* Propagate GPU resources when determining current resources for autoscalingValerij Fredriksen2023-09-281-1/+1
|
* Don't log when throwingJon Bratseth2023-09-051-3/+0
|
* Handle 'wrong order'Jon Bratseth2023-09-051-3/+30
|
* Ignore spurious exception introduced in QuestDb 7.3.1Jon Bratseth2023-09-041-0/+1
|
* Default is no access, so must enable it explicit.Henning Baldersheim2023-09-011-1/+4
|
* Use questdb 7.3.1Henning Baldersheim2023-09-011-5/+6
|
* Get TableToken on demand.Henning Baldersheim2023-08-311-4/+5
|
* questdb 6.7Henning Baldersheim2023-08-311-8/+9
|
* 6.6.1Henning Baldersheim2023-08-311-4/+1
|
* Remove debugValerij Fredriksen2023-08-161-3/+0
|
* Remove debugValerij Fredriksen2023-08-161-2/+0
|
* Adjust by idealJon Bratseth2023-08-161-5/+8
|
* Don't use BCPInfo to scale downJon Bratseth2023-08-161-1/+2
|
* More accurate accounting of redistribution costJon Bratseth2023-08-162-4/+22
|
* Test account for change costJon Bratseth2023-08-151-1/+1
|
* AllocatableClusterresources -> AllocatableResourcesJon Bratseth2023-08-155-67/+65
|
* No functional changes only: Wrap current resources in modelJon Bratseth2023-08-156-68/+69
|
* Consider switching cost when choosing host flavorJon Bratseth2023-08-152-21/+19
|
* Consider switching cost when choosing resourcesJon Bratseth2023-08-154-1/+127
|
* Leave 15% headroom when scaling downJon Bratseth2023-07-311-1/+1
|
* Merge pull request #27897 from vespa-engine/freva/fixMartin Polden2023-07-261-3/+2
|\ | | | | Verify selected flavor is within resource limits
| * Add NodeResourceLimits to NodeRepositoryValerij Fredriksen2023-07-251-3/+2
| |
* | Set aside enough headroom rather than not scaling down at allJon Bratseth2023-07-251-3/+3
| |
* | CleanupJon Bratseth2023-07-252-10/+11
| |
* | Set aside space for headroom in one passJon Bratseth2023-07-253-37/+22
|/
* Don't scale down if we are likely to scale back upJon Bratseth2023-07-202-21/+53
| | | | | | | | | When we decide to scale number of nodes down we'll see a lower ideal load because we need to be able to handle one node going down. This may lead us to be closer to ideal (at current peak load) than otherwise anticipated, such that we are quite likely to soon scale back up. This checks for that and avoids scaling down dimensions where this is the case.
* Remove scaling duration min for content clustersJon Bratseth2023-07-142-1/+1
| | | | This has been a safety measure, which doesn't seem necessary any more.