Docker starts the elasticsearch image and solves the error after mounting the directory

Downloaded an es image from docker hub, version 6.4.2, the details are as follows:

The more important ones are these two. The first is the working directory, and the mount directory also needs to correspond to this; the second is the startup command, where a pre-written startup script is specified. So I started an empty container to check what was going on inside it:

The directory structure inside the container is as shown above. Data is used to store data, and logs is used to store logs.

Next, check the startup script

/usr/local/bin/docker-entrypoint.sh

I only have a limited understanding of the first half, but the last part is really related to the mount directory. It deals with the operations after the mount directory, which roughly means:

If it is the root user (docker starts the container, the default is to start the container as the root user), and the TAKE_FILE_OWNERSHIP variable exists, then change the two directories /usr/share/elasticsearch/{data,logs} to belong to user 1000 (you can also see here that the final data storage path is data, so it should be mounted under data).

User with id 1000:

It is the elasticsearch user, so if you don't mount any directory, you can start the container directly. If you mount it, then add a variable and assign any value, and es can start normally.

docker run -itd -v /root/es-data/:/usr/share/elasticsearch/data -e TAKE_FILE_OWNERSHIP=111 -p 9200:9200 --name es elasticsearch:6.4.2

Supplementary knowledge: Record a docker installation of elasticsearch and the pitfalls encountered

First give a line of command

docker run -d -p 9200:9200 -p 9300:9300 -e "ES_JAVA_OPTS=-Xms512m -Xmx512m" --name=<your es name> elasticsearch:<es version>

Use this command to directly install and run the es docker image container

question

Use the command docker run -d -p 9200:9200 -p 9300:9300 --name=<your es name> elasticsearch:<es version> to start, and then to see if it is started, use the docker ps command to find:

es did not start.

So I used the docker ps -a command again:

Found that the startup was terminated.

Then use the docker logs -f <container id> command to view the es startup log:

Scroll to the bottom:

It looks like this:

Power management:

Memory: 4k page, physical 1882892k(89076k free), swap 0k(0k free)

vm_info: OpenJDK 64-Bit Server VM (25.181-b13) for linux-amd64 JRE (1.8.0_181-8u181-b13-2~deb9u1-b13), built on Oct 22 2018 18:05:23 by "pbuilder" with gcc 6.3.0 20170516

time: Fri Nov 23 07:00:34 2018
elapsed time: 0 seconds (0d 0h 0m 0s)

So I copied the error message and Googled it, and found the problem. The default memory configuration for elasticsearch5.+ is 2g. I only gave docker 2g of memory, so it was GG.

After deleting the old container and the old image, add -e "ES_JAVA_OPTS=-Xms512m -Xmx512m" to the docker startup command to solve the problem

docker ps :

Finally, visit port 9200:

OJ

The above article about how to solve the error after starting the elasticsearch image in docker and mounting the directory is all I want to share with you. I hope it can give you a reference, and I also hope that you will support 123WORDPRESS.COM.