Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Revert "Revert "Enable ZK access control by default"" | Jon Bratseth | 2016-11-10 | 2 | -9/+5 |
| | |||||
* | Merge pull request #1058 from yahoo/bratseth/field-operation-order | gjoranv | 2016-11-10 | 1 | -3/+4 |
|\ | | | | | Order field operations | ||||
| * | Order field operations | Jon Bratseth | 2016-11-10 | 1 | -3/+4 |
| | | |||||
* | | Always recycle failed nodes in CI and CD | Martin Polden | 2016-11-10 | 2 | -8/+29 |
| | | |||||
* | | Test that patching dirty node does not increase reboot generation | Martin Polden | 2016-11-10 | 1 | -0/+22 |
| | | |||||
* | | Make test naming style consistent | Martin Polden | 2016-11-10 | 1 | -8/+8 |
|/ | |||||
* | Only increase wanted reboot generation when node move to dirty from another ↵ | Harald Musum | 2016-11-10 | 1 | -1/+1 |
| | | | | state | ||||
* | Revert "Enable ZK access control by default" | Henning Baldersheim | 2016-11-09 | 2 | -5/+9 |
| | |||||
* | Merge pull request #1043 from yahoo/bratseth/enable-zk-access-control | Jon Bratseth | 2016-11-09 | 2 | -9/+5 |
|\ | | | | | Enable ZK access control by default | ||||
| * | Enable ZK access control by default | Jon Bratseth | 2016-11-09 | 2 | -9/+5 |
| | | |||||
* | | Increase reboot gen. when node is moved to dirty | Martin Polden | 2016-11-09 | 3 | -1/+32 |
| | | |||||
* | | Rewrite test to use ProvisioningTester | Martin Polden | 2016-11-09 | 2 | -50/+31 |
| | | |||||
* | | Remove unused methods | Martin Polden | 2016-11-09 | 1 | -14/+0 |
| | | |||||
* | | More validation | Jon Bratseth | 2016-11-09 | 1 | -2/+14 |
|/ | |||||
* | Use a static variable to communicate | Jon Bratseth | 2016-11-09 | 2 | -16/+15 |
| | | | | | | Some third party code somewhere is calling setProperties, wiping all properties set inside the VM at random times, so communicate through a static variable instead. | ||||
* | Add more logging | Jon Bratseth | 2016-11-08 | 1 | -0/+3 |
| | |||||
* | Merge pull request #1001 from yahoo/mpolden/node-repository-owners | Jon Bratseth | 2016-11-02 | 1 | -0/+1 |
|\ | | | | | Update owners | ||||
| * | Update owners | Martin Polden | 2016-11-02 | 1 | -0/+1 |
| | | |||||
* | | Add temporary logging + whitespace fixes | Jon Bratseth | 2016-11-01 | 2 | -5/+5 |
|/ | |||||
* | Use CI image for local zone | Håkon Hallingstad | 2016-10-28 | 1 | -2/+2 |
| | |||||
* | Remove unnecessary dep | Jon Bratseth | 2016-10-25 | 1 | -6/+0 |
| | |||||
* | Docker image should not be considered when allocating nodes | Harald Musum | 2016-10-25 | 3 | -4/+19 |
| | | | | | | * Ignore (wanted) docker image when allocating nodes, since this is not an attribute of the node as it is used today, it might change at any time (e.g. when upgrading Vespa) | ||||
* | Removed serialized fields that are not used anymore | Harald Musum | 2016-10-24 | 3 | -82/+4 |
| | |||||
* | Clearer OutOfCapacity messages | Jon Bratseth | 2016-10-21 | 1 | -10/+10 |
| | | | | | Dropping the "because we want to retire existing nodes" message as it is confusing at best with the normal cause of this state which is that the user want to change from one flavor to another. | ||||
* | Merge pull request #779 from yahoo/bratseth/node-type-filter | Frode Lundgren | 2016-10-18 | 4 | -3/+49 |
|\ | | | | | Add node type filter | ||||
| * | Add node type filter | Jon Bratseth | 2016-10-04 | 4 | -3/+49 |
| | | |||||
* | | Fix expected value after changes in config model | Harald Musum | 2016-10-17 | 1 | -1/+1 |
| | | |||||
* | | Nonfunctional changes | Jon Bratseth | 2016-10-17 | 1 | -3/+3 |
| | | |||||
* | | log info for regular dev deployments | Kristian Aune | 2016-10-10 | 1 | -1/+1 |
| | | |||||
* | | Merge pull request #786 from ↵ | Harald Musum | 2016-10-06 | 1 | -1/+1 |
|\ \ | | | | | | | | | | | | | yahoo/bratseth/decrease-inactive-timeout-in-non-prod Decrease inactive timeout in non-prod | ||||
| * | | Decrease inactive timeout in non-prod | Jon Bratseth | 2016-10-05 | 1 | -1/+1 |
| | | | |||||
* | | | Uglier code to avoid warnings | Jon Bratseth | 2016-10-05 | 1 | -9/+9 |
| | | | |||||
* | | | Merge pull request #774 from yahoo/bratseth/fail-max-1-node-of-each-type | Martin Polden | 2016-10-05 | 7 | -387/+522 |
|\ \ \ | |/ / |/| | | Bratseth/fail max 1 node of each type | ||||
| * | | Test proxy node failing | Jon Bratseth | 2016-10-04 | 6 | -68/+104 |
| | | | |||||
| * | | Refactor test to avoid chaos and suffering | Jon Bratseth | 2016-10-04 | 3 | -338/+415 |
| | | | |||||
| * | | Fail at most one node of any type!=tenant | Jon Bratseth | 2016-10-04 | 3 | -4/+26 |
| |/ | |||||
* / | Support requiring specific capacity | Jon Bratseth | 2016-10-05 | 1 | -2/+2 |
|/ | |||||
* | Prefer exact match of non-stock nodes | Jon Bratseth | 2016-10-03 | 7 | -46/+85 |
| | |||||
* | Prefer exact flavor if non-canonical | Jon Bratseth | 2016-09-27 | 5 | -19/+70 |
| | |||||
* | Periodically maintain the infrastructure application | Jon Bratseth | 2016-09-21 | 2 | -2/+2 |
| | |||||
* | Test provisioning of proxies | Jon Bratseth | 2016-09-20 | 10 | -14/+136 |
| | |||||
* | Support provision requests by node type | Jon Bratseth | 2016-09-20 | 5 | -52/+157 |
| | |||||
* | Distinguish saturation and sufficiency | Jon Bratseth | 2016-09-20 | 1 | -10/+15 |
| | |||||
* | Refactor/cleanup only | Jon Bratseth | 2016-09-20 | 2 | -15/+7 |
| | |||||
* | Allow nodes to be requested by type | Jon Bratseth | 2016-09-20 | 28 | -174/+187 |
| | |||||
* | Use Node.State from node repo instead of duplicating value in our own class | Harald Musum | 2016-09-19 | 3 | -6/+6 |
| | |||||
* | Revert "Revert "Remove /hack/provision legacy API"" | Harald Musum | 2016-09-16 | 8 | -450/+0 |
| | |||||
* | Revert "Remove /hack/provision legacy API" | Harald Musum | 2016-09-16 | 8 | -0/+450 |
| | |||||
* | Remove /hack/provision legacy API | Harald Musum | 2016-09-15 | 8 | -450/+0 |
| | |||||
* | Use an enum for flavor type | Harald Musum | 2016-09-12 | 26 | -57/+56 |
| |