Relumin (Redis cluster admin)

What's Relumin ?

Relumin is a REdis cLUster adMIN tool.

Main feature is ...

• Visualize Redis cluster status
• Operating Cluster
  • Add node, Resharding, Delete node, Replicate, Failover...etc
• Visualize Redis node metrics as chart
  • Metrics is get by http://redis.io/commands/info
• Alert
  • You can set threshold freely.

Why develop ?

• There are some redis tool, but they don't support Redis cluster.
• It is difficult to understand cluster status. I want visualize.
• Other cluster storages have GUI tool.
  • For example,
    • Riak has "Riak control(http://docs.basho.com/riak/latest/ops/advanced/riak-control/)".
    • Cassandra has "Datastax(http://www.datastax.com/)".

So I developed.

Screenshot

Cluster status

Metics chart

Setting alert

Getting start

Very easy.

1. Download JAR file

Please download from this.
https://github.com/be-hase/relumin/releases

2. Install Java8 on your server

Install Java8 on your server.
( I tested Oracle JDK. )

3. Install Redis(not cluster) on your server

Install Redis on your server.

Redis is used for store meta data and metrics data.

Memory is expensive. So I will support MySQL. Prease wait...

4. Write "relumin.yml" (config file)

Please write config file.
Specify this config path when you start Relumin.

Example:

host: 10.10.10.10

server:
  port: 8080
  monitorPort: 20080

redis:
  host: localhost
  port: 10000

auth:
  enabled: true
  allowAnonymous: true

notice:
  mail:
    host: smtp.example.com
    port: 25
    from: relumin-alert@example.com

scheduler:
  collectStaticsInfoMaxCount: 10000

This is property table.

property name	default	required	description
host			Host or IP of server which Relumin running on. This is used for alert mail.
server.port	8080		Port of Relumin server.
server.monitorPort	20080		Port of Relumin server for monitoring.
redis.prefixKey	_relumin		Prefix key of Redis. This prefix key is used for meta data and metrics data.
redis.host		Y	Redis's host.
redis.port		Y	Redis's port
auth.enabled	false		If this value is false, Every user can execute all operation. (Cluster operation, User CRUD, Notification CRUD...etc)
auth.allowAnonymous	false		If this value is true, anonymous user can access Relumin. But only view.
notice.mail.host			Host of SMTP server.
notice.mail.port			Port of SMTP server.
notice.mail.user			User of SMTP Auth.
notice.mail.password			Password of SMTP Auth.
notice.mail.charset	UTF-8		Charset of alert mail.
notice.mail.from			"From" mail address. This value can be overriden by cluster setting.
scheduler.refreshClustersIntervalMillis	120000		Interval(milliseconds) of refresh meta data of Redis cluster. Recommend default.
scheduler.collectStaticsInfoIntervalMillis	120000		Interval(milliseconds) of collect metrics data. If threshold condition is satisfied, send notice.
scheduler.collectStaticsInfoMaxCount	1500		Save count of metrics data. If you want save latest 1 week, you specify "5040". (60 * 24 * 7) / 2 = 5040
scheduler.collectSlowLogMaxCount	1500		Save count of slowlog.
outputMetrics.fluentd.enabled	false		If you want to forward metris data into fluentd, specify true.
outputMetrics.fluentd.host			Host of Fluentd.
outputMetrics.fluentd.port			Port of Fluentd.
outputMetrics.fluentd.timeout	3000		Timeout of Fluentd Java Library.
outputMetrics.fluentd.bufferCapacity	1048576		BufferCapacity of Fluentd Java Library.
outputMetrics.fluentd.tag	relumin		Tag of Fluentd. Send by below format. <outputMetrics.fluentd.tag>.outputMetrics.fluentd.nodeTag>
outputMetrics.fluentd.nodeTag	node		Ref avobe.

5. Start Jar (FINISH !!)

Please start jar file.

You can specify log parameter.

• log.type : console or file
• log.level : DEBUG, INFO, WARN, ERROR... (logback's level)
• log.dir : /path/your/logDir

java -Dconfig=/path/your/config/relumin.yml -Dlog.type=file -Dlog.level=INFO -Dlog.dir=/path/your/logDir -jar relumin-0.0.1.jar > /dev/null 2>&1 &

Relumin run on JVM, so you can specify JVM option. Like a below.

JVM_OPT="-server -Dfile.encoding=UTF-8 -Xms2g -Xmx2g \
-XX:SurvivorRatio=8 -XX:TargetSurvivorRatio=80 \
-XX:+UseConcMarkSweepGC -XX:+CMSParallelRemarkEnabled \
-XX:CMSInitiatingOccupancyFraction=70 \
-XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=${LOG_PATH}/relumin -verbose:gc \
-XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:${LOG_PATH}/relumin/relumin.gc.log \
-XX:+UseGCLogFileRotation -XX:GCLogFileSize=512m -XX:NumberOfGCLogFiles=5"

java -Dconfig=/path/your/config/relumin.yml -Dlog.type=file -Dlog.level=INFO -Dlog.dir=/path/your/logDir -jar relumin-0.0.1.jar > /dev/null 2>&1 &

Authenticate and Authorize (Option)

Relumin support role-based authenticate&authorize.

Role

Relumin user has 2 role.

role name	description
VIEWER	Only view user.
RELUMIN_ADMIN	This user execute all Relumin operation. (Cluster operation, User CRUD, Notification CRUD...etc)

About adding user

A user who is RELUMIN_ADMIN can add user from 'http:///#/users'.

How do I add First RELUMIN_ADMIN user?
Please specify 'auth.enabled' false, and add RELUMIN_USER, and 'auth.enabled' true.

Anonymous user

You can configure whether anonymous user can access Relumin or not by specifying 'auth.allowAnonymous'. If this value is true, anonymous user can use relumin as VIEWER role.

Contribute

For local developing

1. Build docker image. docker build -t relumin/relumin-test .
2. Run docker. docker run -d --net="host" -p 9000:9000 -p 10000:10000 -p 10001:10001 -p 10002:10002 -p 10003:10003 -p 10004:10004 -p 10005:10005 -p 20000:20000 -p 20001:20001 -p 20002:20002 -p 20003:20003 -p 20004:20004 -p 20005:20005 -p 10010:10010 -p 10011:10011 -p 10012:10012 -p 10013:10013 -p 10014:10014 -p 10015:10015 relumin/relumin-test
3. Copy config files. cp src/main/resources/relumin-local-conf.yml.sample src/main/resources/relumin-local-conf.yml, cp src/test/resources/application.yml.sample src/test/resources/application.yml
4. Replace "YOUR_DOCKER_HOST" of config files.
5. Write your code.
6. Run unit test. mvn clean test
7. Run spring boot app. mvn clean spring-boot:run -Dconfig=relumin-local-conf.yml

Milestone (Maybe...)

• Support share URL
• Support MySQL for meta data and metrics data.

Appendix

Architecture

Server

• Written by Java.
• Based on Spring Boot.

Front

• Single Page Application.
• Written by pure javascript.
• Based on React(Flux), Webpack and Gulp.