Lobster is a distributed Kubernetes container logging system that can be installed in a Kubernetes cluster to store and query logs from Kubernetes containers.
At Kubernetes, we need to consider cluster-level logging architectures.
The main thing is the need for storage with different life cycles.
- Need to look up the container logs that the Pod had before it was relocated
- Need to look up the rotated logs
- Need to look up by any groups or time ranges
The main concepts of Lobster are:
- Supports logs for all containers in Kubernetes that are rotated and relocated
- Supports the ability to view multiple logs from a single endpoint based on time and Kubernetes objects
- Supports easy installation using the node disk without having to build separate storage for the log system in the Kubernetes environment
For more information, please refer to the documents below.
I'm preparing public images to make it easier to use lobster.
Lobster store
tails and collects the log files with the container'sstdout/stderr
in container log directories- If the container is mounting an emptydir volumes,
Lobster store
tails and collects the log files in emptydir volume directories. There are some requirements to tail files as below- Log line should start with
{RFC 3339 timestamp}
- Log file should have
.log
extension
- Log line should start with
- Search logs from multiple distributed clusters in one view
- Search logs with units of Kubernetes objects(e.g. Namespace, Label, Set, Pod, and Container)
- Supports log search options to include/exclude logs based on google re2
- Supports logs sink custom resources
Log metric
: Create a metric by defining a pattern of log lines within a log sink(e.g. countGET
logs)Log export
: Export logs by defining a pattern of log lines within a log sink(e.g. sendGET
logs to an external bucket)
- To prevent overload on nodes caused by excessive container logs, there are configurable limits on the number of logs per container
- 1MB/s ~ 30MB/s
- 30k lines/s
- To manage the capabilities of node disks, there are configurable limits on log storage retention
- 1 week, 2 GB retention
- Supports web page view from query component
Lobster store
stores logs produced from containers and provides APIs to query themLobster query
can perform log queries on multipleLobster stores
. The query can be done on APIs or the Lobster web page. The query is fanned out to eachLobster store
andLobster query
responds by aggregating the query results
See more in Lobster store and Lobster query documents.
Lobster global query
requests to theLobster query
of each cluster to obtain the chunks containing an address ofLobster store
Lobster global query
queries eachLobster store
directly based on the chunks and it responds by aggregating the query results
See more in Lobster query document.
It goes a bit further from Multi-clusters model
and supports Log sink
Log sink
is the action of exporting logs to external storage or creating metricsLobster operator
definesLog sink custom resources
and manages settings for log export/metricLobster matcher
produces metrics for matching logs based onLog sink (log metric)
rules. This metrics follow the prometheus data modelLobster exporter
posts matching logs to external storage based onLog sink (log export)
rules
See more in Log sink document.
Lobster
Copyright (c) 2024-present NAVER Corp.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.