Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Revert "Revert "Revert "Revert "Create a resourcepool so that we do not need ↵ | Henning Baldersheim | 2020-02-19 | 1 | -4/+4 |
| | | | | to reconnect to content …"""" | ||||
* | Revert "Revert "Revert "Create a resourcepool so that we do not need to ↵ | Harald Musum | 2020-02-19 | 1 | -4/+4 |
| | | | | reconnect to content …""" | ||||
* | Revert "Revert "Create a resourcepool so that we do not need to reconnect to ↵ | Henning Baldersheim | 2020-02-14 | 1 | -4/+4 |
| | | | | content …"" | ||||
* | Revert "Create a resourcepool so that we do not need to reconnect to content ↵ | Harald Musum | 2020-02-14 | 1 | -4/+4 |
| | | | | …" | ||||
* | Create a resourcepool so that we do not need to reconnect to content cluster ↵ | Henning Baldersheim | 2020-02-13 | 1 | -4/+4 |
| | | | | on changes to container cluster. | ||||
* | Move pingfactory to constructor. | Henning Baldersheim | 2020-02-04 | 1 | -4/+4 |
| | |||||
* | - Shut down monitoring thread. | Henning Baldersheim | 2019-10-04 | 1 | -9/+9 |
| | | | | | - Remove fs4 cleanup. - Add some more debug information for group status. | ||||
* | Feature flag to enable protobuf in search protocol as default; protobuf ping | Olli Virtanen | 2019-03-28 | 1 | -4/+4 |
| | |||||
* | Adaptive load balancing scheduler, coverage calculation fixes | Olli Virtanen | 2018-12-19 | 1 | -25/+107 |
| | |||||
* | Adaptive timeout support in java dispatch | Olli Virtanen | 2018-11-23 | 1 | -14/+14 |
| | |||||
* | Merge pull request #7679 from ↵ | Jon Bratseth | 2018-11-18 | 1 | -9/+8 |
|\ | | | | | | | | | vespa-engine/ollivir/group-affinity-with-searchpath Java dispatcher group affinity through searchPath | ||||
| * | Group affinity through searchPath; remove Query from LB assignment | Olli Virtanen | 2018-11-16 | 1 | -9/+8 |
| | | |||||
* | | When pooling ClusterMonitors and SearchClusters you will add a cluster every ↵ | Henning Baldersheim | 2018-11-15 | 1 | -5/+5 |
|/ | | | | | | | | time there is a reconfig. As nothing will purge them, you both keep stuff alive forever and end up with more clusters that you have. Hence the magic for not removing vipstatus when there are multiple clusters kick in preventing nodes being taken OOR. Now it is using the ComponentId for identifying a cluster. | ||||
* | Java dispatcher refactoring and cluster monitoring improvements | Olli Virtanen | 2018-11-02 | 1 | -21/+22 |
| | |||||
* | Pull more configuration parameters from config model to internal dispatcher ↵ | Olli Virtanen | 2018-10-24 | 1 | -10/+10 |
| | | | | config | ||||
* | Remove group affinity - conflicts with profiles | Olli Virtanen | 2018-10-12 | 1 | -21/+0 |
| | |||||
* | Multiple node java dispatch support | Olli Virtanen | 2018-09-24 | 1 | -13/+2 |
| | |||||
* | Reuse node group for allocations by same query | Olli Virtanen | 2018-09-14 | 1 | -9/+31 |
| | |||||
* | Added copyrights and javadocs. Logic and naming in LB cleaned up | Olli Virtanen | 2018-09-06 | 1 | -9/+12 |
| | |||||
* | Java dispatch support for multiple groups of single nodes | Olli Virtanen | 2018-09-06 | 1 | -4/+54 |
| | |||||
* | Create code path for extending java side dispatcher | Olli Virtanen | 2018-08-30 | 1 | -0/+64 |