Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Remove unused config field | Harald Musum | 2022-09-06 | 1 | -1/+0 |
| | |||||
* | Make FleetControllerOptions immutable and support builder pattern | Harald Musum | 2022-08-31 | 3 | -63/+56 |
| | |||||
* | Remove useless test that requires hacks in other places | Harald Musum | 2022-08-22 | 1 | -70/+0 |
| | | | | | Want to require zooKeeperServerAddress and slobrokConnectionSpecs being non-null | ||||
* | Revert "Cluster controller unit test cleanup, part 4 [run-systemtest]" | Harald Musum | 2022-08-22 | 1 | -0/+70 |
| | |||||
* | Remove useless test that requires hacks in other places | Harald Musum | 2022-08-17 | 1 | -70/+0 |
| | | | | | Want to require zooKeeperServerAddress and slobrokConnectionSpecs being non-null | ||||
* | Convert clustercontroller-apps to junit5 | Bjørn Christian Seime | 2022-07-29 | 7 | -51/+60 |
| | |||||
* | Set project version to 8-SNAPSHOT | gjoranv | 2022-06-08 | 1 | -2/+2 |
| | |||||
* | install_jar CMake function | Håkon Hallingstad | 2022-05-20 | 1 | -1/+1 |
| | |||||
* | Use '@Inject' from 'annotations' in multiple bundles | Bjørn Christian Seime | 2022-05-06 | 4 | -4/+4 |
| | |||||
* | Shut down ClusterController from last configurer, thus before ZK | Jon Marius Venstad | 2022-03-11 | 4 | -20/+52 |
| | |||||
* | Revert "Ensure ClusterController shuts down before ZK cluster is uses" | Arnstein Ressem | 2022-03-11 | 1 | -7/+1 |
| | |||||
* | Ensure ClusterController shuts down before ZK cluster is uses | Jon Marius Venstad | 2022-03-11 | 1 | -1/+7 |
| | |||||
* | Extend ThreadedHttpReqHndlr instead of deprecated LoggingReqHndlr | gjoranv | 2022-01-17 | 3 | -5/+5 |
| | |||||
* | Remove config generation -1/0 from CC at :19050/status/<clustername>/config | Håkon Hallingstad | 2021-10-20 | 1 | -1/+1 |
| | |||||
* | Update Verizon Media copyright notices. | gjoranv | 2021-10-07 | 3 | -3/+3 |
| | |||||
* | Update 2017 copyright notices. | gjoranv | 2021-10-07 | 12 | -12/+12 |
| | |||||
* | Revert "Avoid safe mutations in master moratorium and increase first cluster ↵ | Håkon Hallingstad | 2021-03-24 | 1 | -14/+3 |
| | | | | state broadcast deadline [run-systemtest]" | ||||
* | Increase the minimum time before first cluster state broadcast [run-systemtest] | Håkon Hallingstad | 2021-03-19 | 1 | -3/+14 |
| | |||||
* | High graph constructino timeout | Jon Marius Venstad | 2021-02-20 | 1 | -1/+1 |
| | |||||
* | Die after 1 minute if no ZK-connection when constructing CC | Jon Marius Venstad | 2021-02-19 | 1 | -1/+3 |
| | |||||
* | Merge pull request #16527 from ↵ | Jon Marius Venstad | 2021-02-16 | 3 | -11/+19 |
|\ | | | | | | | | | vespa-engine/jonmv/create-only-one-cluster-controller Avoid recreation of ClusterController when config changes | ||||
| * | Avoid recreation of ClusterController when config changes | Jon Marius Venstad | 2021-02-16 | 3 | -11/+19 |
| | | |||||
* | | Also shut down Curator instances created during clustercontroller component ↵ | Jon Marius Venstad | 2021-02-14 | 1 | -2/+3 |
|/ | | | | setup | ||||
* | Support configurable feed block hysteresis on the cluster controller | Tor Brede Vekterli | 2021-02-10 | 2 | -1/+4 |
| | | | | | | | | | | | | | | | | | | | Adds an absolute number delta that is subtracted from the feed block limit when a node has a resource already in feed blocked state. This means that there's a lower watermark threshold that must be crossed before feeding can be unblocked. Avoids flip-flopping between block states. Default is currently 0.0, i.e. effectively disabled. To be modified later for system tests and trial roll-outs. A couple of caveats with the current implementation: * The cluster state is not recomputed automatically when just the hysteresis threshold is crossed, so the description will be out of date on the content nodes. However, if any other feed block event happens (or the hysteresis threshold is crossed), the state will be recomputed as expected. This does not affect correctness, since the feed is still to be blocked. * A node event remove/add pair is emitted for feed block status when the hysteresis threshold is crossed and there's a cluster state recomputation. | ||||
* | Fix import | Harald Musum | 2021-02-09 | 1 | -1/+1 |
| | |||||
* | Cleanup: Remove unnecessary and unused methods, simplify | Harald Musum | 2021-02-08 | 2 | -17/+24 |
| | |||||
* | Add cluster feed block support to cluster controller | Tor Brede Vekterli | 2021-01-27 | 2 | -1/+10 |
| | | | | | | | | | | | | | | | Will push out a new cluster state bundle indicating cluster feed blocked if one or more nodes in the cluster has one or more resources exhausted. Similarly, a new state will be pushed out once no nodes have resources exhausted any more. The feed block description currently contains up to 3 separate exhausted resources, possibly across multiple nodes. A cluster-level event is emitted for both the block and unblock edges. No hysteresis is present yet, so if a node is oscillating around a block-limit, so will the cluster state. | ||||
* | Add back javadoc | Harald Musum | 2020-12-16 | 1 | -0/+4 |
| | |||||
* | Inject VespaZooKeeperServer in constructor | Harald Musum | 2020-12-16 | 1 | -1/+7 |
| | |||||
* | Revert "Revert " Stop depending on ZooKeeperProvider, take 2"" | Harald Musum | 2020-12-16 | 4 | -48/+2 |
| | |||||
* | Revert " Stop depending on ZooKeeperProvider, take 2" | Harald Musum | 2020-12-16 | 4 | -2/+48 |
| | |||||
* | Constructor needs to be public | Harald Musum | 2020-12-15 | 1 | -1/+1 |
| | |||||
* | Revert "Revert "Stop depending on ZooKeeperProvider"" | Harald Musum | 2020-12-15 | 4 | -47/+1 |
| | |||||
* | Revert "Stop depending on ZooKeeperProvider" | Harald Musum | 2020-12-15 | 4 | -1/+47 |
| | |||||
* | Stop depending on ZooKeeperProvider | Harald Musum | 2020-12-15 | 4 | -47/+1 |
| | | | | | | | | VespaZooKeeperServer is enough, ZooKeeperProvider is just an unnecessary extra layer. In addition neither provides any guarantee that the the server has started and is working. ClusterController has code that verifies that connecting to zookeeper works, that should be sufficient. | ||||
* | Merge branch 'master' into ↵ | Harald Musum | 2020-12-14 | 1 | -12/+13 |
|\ | | | | | | | hmusum/disallow-clustercontroller-with-no-zookeeper-cluster | ||||
| * | Use CuratorFramework.blockUntilConnected() to verify connection | Harald Musum | 2020-12-14 | 1 | -6/+2 |
| | | |||||
| * | Refactor code and document code to verify that ZooKeeper is working | Harald Musum | 2020-12-14 | 1 | -12/+17 |
| | | |||||
* | | Remove unnecessary argument | Harald Musum | 2020-12-14 | 3 | -18/+18 |
| | | |||||
* | | Disallow creating a cluster controller having no zookeeper cluster | Harald Musum | 2020-12-14 | 2 | -28/+8 |
|/ | |||||
* | Revert "Remove code that nobody understands the need for" | Arnstein Ressem | 2020-12-09 | 1 | -2/+11 |
| | |||||
* | Remove code that nobody undrstands the need for | Harald Musum | 2020-12-08 | 1 | -11/+2 |
| | | | | | | Cannot find any usage of nodes created. This might be some initialization code to check if operations work, but I cannot see the need for it. If this breaks something we can at least document why this is needed. | ||||
* | Move code in clustercontroller-apputils into clustercontroller-apps | Harald Musum | 2020-12-01 | 6 | -6/+278 |
| | | | | | Code in clustercontroller-apputils is now only used from clustercontroller-apps, so those two modules can be merged | ||||
* | Rename method | Harald Musum | 2020-11-18 | 1 | -1/+1 |
| | |||||
* | LogLevel -> Level for isLoggable() | gjoranv | 2020-04-25 | 1 | -1/+1 |
| | |||||
* | Import java.util.logging.Level instead of com.yahoo.log.LogLevel | gjoranv | 2020-04-25 | 1 | -1/+1 |
| | |||||
* | Remove use-bucket-space-metric feature flag | Håkon Hallingstad | 2020-01-26 | 1 | -1/+0 |
| | | | | | | | | | | The flag controlled config read by the Cluster Controller. Therefore, I have left the ModelContextImpl.Properties method and implementation (now always returning true), but the model has stopped using that method internally, and the config is no longer used in the CC. The field in the fleetcontroller.def is left unchanged and documented as deprecated. | ||||
* | Use bucket_space metric in retirement | Håkon Hallingstad | 2020-01-17 | 1 | -0/+1 |
| | | | | | | | | | | | | This makes the Cluster Controller use the vds.datastored.bucket_space.buckets_total, dimension bucketSpace=default, to determine whether a content node manages zero buckets, and if so, will allow the node to go permanently down. This is used when a node is retiring, and it is to be removed from the application. The change is guarded by the use-bucket-space-metric, default true. If the new metric doesn't work as expected, we can revert to using the current/old metric by flipping the flag. The flag can be controlled per application. | ||||
* | Revert "Revert "Reapply "Create seperate bundles compiled against different ↵ | Harald Musum | 2019-10-28 | 1 | -1/+1 |
| | | | | ZooKeeper versions""" | ||||
* | Revert "Reapply "Create seperate bundles compiled against different ↵ | Harald Musum | 2019-10-28 | 1 | -1/+1 |
| | | | | ZooKeeper versions"" |