aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* There's only values in this metric typemetrics-jmx-for-brokersStaffan Olsson2018-02-031-1/+1
|
* At first i suspected that order matters, but this also worksStaffan Olsson2018-02-031-2/+2
|
* ”In a production Kafka cluster, an offline partitionStaffan Olsson2018-02-031-1/+2
| | | | | | | | | | may be impacting the producer clients, losing messages or causing back-pressure in the application. This is most often a “site down” type of problem and will need to be addressed immediately.” Excerpt from: Neha Narkhede, Gwen Shapira, and Todd Palino. ”Kafka: The Definitive Guide”. We now export kafka_controller_kafkacontroller_value{name="OfflinePartitionsCount",} and friends. See #140 for why.
* It was the liveness probe that killed the metrics container,Staffan Olsson2018-01-291-2/+2
| | | | | | and we might not need liveness if we have alerts for stale metrics. This reverts commit f1e6e96231465775ffa0ded5b745774182fec5ff.
* Removes liveness probes, to focus on memory limits,Staffan Olsson2018-01-191-5/+0
| | | | | | and rely on metric staleness alerts instead for exporter liveness. This reverts commit 74a5177270301f3cbe0342657d8bcf70d1ae76ed.
* With 120M you get OOMKilled even on brokers with modest metrics volumesStaffan Olsson2018-01-191-2/+2
| | | | This reverts commit 66255af447373f9b14e338a4b819dddad483a257.
* With kubectl patch, this folder kan be scoped as Prometheus supportStaffan Olsson2018-01-192-1/+1
|
* On start the metrics pod gesses that it has ~45%Staffan Olsson2018-01-191-2/+2
| | | | | | of the memory limit as "Max. Heap Size (Estimated)" Reducing limits as experiment for #112.
* Evaluates JVM memory limit awarenessStaffan Olsson2018-01-191-2/+4
| | | | | Interesting input for #112, for use with broker and zk pods in addition to KAFKA_HEAP_OPTS.
* This is an optional feature, but belongs to the broker pods,Staffan Olsson2018-01-191-0/+2
| | | | so let's evaluate kubectl patch to add it
* Adds liveness probe from the metrics-improve-scrape-times branchStaffan Olsson2018-01-191-0/+5
|
* Upgrades jmx exporter to 0.2.0Staffan Olsson2018-01-191-1/+1
|
* The metrics part of #49Staffan Olsson2018-01-191-0/+38
|
* Current config from the metrics-improve-scrape-times branch,Staffan Olsson2018-01-191-0/+44
| | | | see https://github.com/Yolean/kubernetes-kafka/pull/49
* Merge pull request #126 from StevenACoffman/docker-k8ssolsson2018-01-182-0/+12
|\ | | | | Add Docker-for-mac k8s specific provisioners for StorageClass
| * docker-k8s: Back to RetainSteve Coffman2018-01-182-2/+2
| |
| * docker-k8s: Allow docker edge kubernetes integration to workSteve Coffman2018-01-182-0/+12
|/
* Removes a test script from back when zookeeper was a separate imageStaffan Olsson2018-01-091-19/+0
| | | | closes #119. Tests in this repo should be dockerized. K8s-ized actually.
* Merge pull request #117 from Yolean/broker-init-pod-lablersolsson2018-01-083-7/+49
|\ | | | | Fix RBAC, set useful labels on broker pods from init script
| * Sets all labels at the end of the init script,broker-init-pod-lablerStaffan Olsson2018-01-082-6/+8
| | | | | | | | | | | | to speed up a bit and reduce the risk that it interferes with actual config. Also if the init script goes crash looping you won't get warnings about labels already being set.
| * Sets other useful labels, for #78 and #56Staffan Olsson2018-01-082-7/+9
| |
| * Revert "Deprecates our own statefulset pod label kafka-broker-id"Staffan Olsson2018-01-082-3/+0
| | | | | | | | This reverts commit 096e3bdfba31d2918520011886a6a0d1f1b08c84.
| * Adds the required RBAC for the init script to set the kafka-broker-id labelStaffan Olsson2018-01-081-0/+41
|/
* Merge pull request #115 from StevenACoffman/cleanupsolsson2018-01-084-7/+5
|\ | | | | Minor cleanups
| * cleanup: Minor cleanupsSteve Coffman2018-01-054-7/+5
|/
* Merge pull request #92 from Yolean/1.8-kube-eventssolsson2017-12-183-0/+170
|\ | | | | Stream kube events to topic, as opt-in folder instead of addon
| * Modernize based on1.8-kube-eventsStaffan Olsson2017-12-183-38/+6
| | | | | | | | | | | | | | * https://github.com/Yolean/kubernetes-kafka/pull/107 * https://github.com/Yolean/kubernetes-kafka/issues/101 * https://github.com/Yolean/fluent-bit-kubernetes-kafka/commit/e045ef1ef03f9fd299575a7aa9ca312e6687e441 * https://github.com/Yolean/kubernetes-kafka/pull/52
| * Uses kafkacat build as in tests and logs-streamingStaffan Olsson2017-11-091-1/+1
| |
| * Fixes manifest. I must have had a local edit last time.Staffan Olsson2017-10-301-1/+1
| |
| * Merge pull request #86 from Yolean/atamon-patch-1solsson2017-10-301-2/+2
| |\ | | | | | | Fix retention from 0.8 days to 8 days
| | * Fix retention from 0.8 days to 8 daysatamon-patch-1Anton Lindgren2017-10-301-2/+2
| |/ | | | | | | | | Silly change, but 0.8 days is very short. Google helped me verify the math.
| * Aligns RBAC with addons->folders for v3.0.0Staffan Olsson2017-10-271-0/+0
| |
| * Fixes JSON syntax, topic is jq:able nowStaffan Olsson2017-10-271-2/+2
| |
| * Needs to be tweaked depending on the churnStaffan Olsson2017-10-271-1/+1
| |
| * An actual assertionStaffan Olsson2017-10-271-10/+72
| |
| * Manifests updated to 1.8 and v3.0.0 pre-mergeStaffan Olsson2017-10-273-2/+10
| |
| * kafkacat's debug output solves the silence, but ...Staffan Olsson2017-10-261-2/+0
| | | | | | | | | | | | to see the error message (like missing RBAC) you basically have to run the curl command. I'd like startup to be as fast as possible, to not miss events, so I hesitate to add checks.
| * bash -e is important, in combination with curl -f, so ...Staffan Olsson2017-10-261-12/+2
| | | | | | | | | | we don't send error messages to the topic (even if it's json -- format is different and it's pretty printed)
| * Argh, curl will pretty print the json ...Staffan Olsson2017-10-261-1/+11
| | | | | | | | and kafkacat will produce each line separately
| * Makes things a bit less obscure if for example RBAC failsStaffan Olsson2017-10-261-2/+12
| |
| * Use an image with new kafkacat and curl versionsStaffan Olsson2017-10-261-1/+1
| |
| * Not a proper test yet, but the PR documents how to use it to see eventsStaffan Olsson2017-08-051-7/+2
| |
| * Moves the example comsumer to the test namespaceStaffan Olsson2017-08-051-5/+4
| |
| * Adds RBAC policy for curl, default service accountStaffan Olsson2017-08-051-0/+30
| |
| * Latest build suppors snappy compressionStaffan Olsson2017-08-052-2/+2
| |
| * Works with my current kafka image, and the test consumer still sees JSONStaffan Olsson2017-08-051-1/+1
| |
| * Testing compression but solsson/kafka:0.11.0.0 throws ...Staffan Olsson2017-08-051-0/+1
| | | | | | | | | | | | | | | | | | Caused by: java.lang.NoClassDefFoundError: Could not initialize class org.xerial.snappy.Snappy at org.xerial.snappy.SnappyInputStream.readFully(SnappyInputStream.java:145) at org.xerial.snappy.SnappyInputStream.readHeader(SnappyInputStream.java:99) at org.xerial.snappy.SnappyInputStream.<init>(SnappyInputStream.java:59) at org.apache.kafka.common.record.CompressionType$3.wrapForInput(CompressionType.java:82) ... 40 more
| * Produces proper json. Let's see when curl gives up and the container restarts.Staffan Olsson2017-08-051-2/+1
| |
| * Quite interesting way to test a command from yaml - see it in logsStaffan Olsson2017-08-051-1/+3
| |
| * wipStaffan Olsson2017-08-051-5/+6
| |
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-05 07:52:17 +0000