ERR_250_ENTRY_ALREADY_EXISTS ou=system already exists

[shashanksheela]

we have a docker-compose.yml will some services in which we are using this image for one service as below
version: '3.6'
services:
ldap:
image: openmicroscopy/apacheds
ports:
- "10389:10389"
- "10636:10636"
- "60088:60088"
- "60464:60464"
- "9090:8080"
- "8443:8443"
volumes:
- ./myconfig.ldif:/bootstrap/conf/config.ldif:ro
- myconfigdata:/var/lib/apacheds
//other services//
volumes:
myconfigdata:

The first time it starts perfectly but later on, it gives ERR_250_ENTRY_ALREADY_EXISTS ou=system already exists and gets terminated. Kindly help us.

full logs
{"log":"Cleaning up /var/lib/apacheds/default/run/apacheds-default.pid\n","stream":"stdout","time":"2018-06-21T07:38:21.869608788Z"}
{"log":"Starting ApacheDS - default...\n","stream":"stdout","time":"2018-06-21T07:38:24.91363281Z"}
{"log":"[07:39:43] ERROR [org.apache.directory.server.wrapper.ApacheDsTanukiWrapper] - Failed to start the service.\n","stream":"stdout","time":"2018-06-21T07:39:43.894071776Z"}
{"log":"org.apache.directory.api.ldap.model.exception.LdapEntryAlreadyExistsException: ERR_250_ENTRY_ALREADY_EXISTS ou=system already exists!\n","stream":"stdout","time":"2018-06-21T07:39:43.894094076Z"}
{"log":"\u0009at org.apache.directory.server.core.partition.impl.btree.AbstractBTreePartition.add(AbstractBTreePartition.java:698)\n","stream":"stdout","time":"2018-06-21T07:39:43.894106484Z"}
{"log":"\u0009at org.apache.directory.server.core.DefaultDirectoryService.initializeSystemPartition(DefaultDirectoryService.java:1764)\n","stream":"stdout","time":"2018-06-21T07:39:43.894110423Z"}
{"log":"\u0009at org.apache.directory.server.core.DefaultDirectoryService.initialize(DefaultDirectoryService.java:1824)\n","stream":"stdout","time":"2018-06-21T07:39:43.894113929Z"}
{"log":"\u0009at org.apache.directory.server.core.DefaultDirectoryService.startup(DefaultDirectoryService.java:1251)\n","stream":"stdout","time":"2018-06-21T07:39:43.89411724Z"}
{"log":"\u0009at org.apache.directory.server.ApacheDsService.initDirectoryService(ApacheDsService.java:364)\n","stream":"stdout","time":"2018-06-21T07:39:43.89412127Z"}
{"log":"\u0009at org.apache.directory.server.ApacheDsService.start(ApacheDsService.java:197)\n","stream":"stdout","time":"2018-06-21T07:39:43.894125062Z"}
{"log":"\u0009at org.apache.directory.server.ApacheDsService.start(ApacheDsService.java:153)\n","stream":"stdout","time":"2018-06-21T07:39:43.894128718Z"}
{"log":"\u0009at org.apache.directory.server.wrapper.ApacheDsTanukiWrapper.start(ApacheDsTanukiWrapper.java:175)\n","stream":"stdout","time":"2018-06-21T07:39:43.894146597Z"}
{"log":"\u0009at org.tanukisoftware.wrapper.WrapperManager$12.run(WrapperManager.java:2788)\n","stream":"stdout","time":"2018-06-21T07:39:43.894150444Z"}

[shashanksheela]

We have tried by setting APACHEDS_INSTANCE still, the issue persists
environment:
APACHEDS_INSTANCE: "myinstance"

[joshmoore]

Hi @shashanksheela . Thanks for the compose example. I may end up adding that to the repo if you don't mind. I'm currently testing with:

version: "3.3"
services:
  ldap:
    image: "openmicroscopy/apacheds:0.6.0"
    ports:
      - "10389:10389"
      - "10636:10636"
      - "60088:60088"
      - "60464:60464"
      - "9090:8080"
      - "8443:8443"
    volumes:
      - ./myconfigdata:/var/lib/apacheds
      # ./myconfig.ldif:/bootstrap/conf/config.ldif:ro

since I don't have your config file. However, this is re-starting fine for me:

$ docker-compose up
Starting ldap_ldap_1 ...
Starting ldap_ldap_1 ... done
Attaching to ldap_ldap_1
ldap_1  | Starting ApacheDS - default...
ldap_1  | [09:31:44] WARN [org.apache.directory.server.core.DefaultDirectoryService] - You didn't change the admin password of directory service instance 'default'.  Please update the admin password as soon as possible to prevent a possible security breach.
^CGracefully stopping... (press Ctrl+C again to force)
Stopping ldap_ldap_1 ... done

Is your LDIF file something you could share with me so I could test?

[joshmoore]

Hmmm.... I'm not sure what's going on here, @shashanksheela. It's working for me:

• with/without your LDIF file
• with the volume set to myconfigdata as you have it or in the current directory like ./myconfigdata

Can you show me a full terminal script of starting a new volume, stopping it, and restarting it, all with the logs?

Edit: by the way:

$ docker --version
Docker version 18.03.1-ce, build 9ee9f40
$ docker-compose --version
docker-compose version 1.17.1, build 6d101fb

[joshmoore]

The only thing thing I can think of right now is that it's due to the filesystem settings somehow, i.e. that the final state of the server isn't being flushed to disk.

[shashanksheela]

Would changing dockerfile accordingly helps?

[joshmoore]

If so, I don't know how. My work in #12 was focused specifically on fixing this restart issue, but for some reasons it's still happening in your environment.

[shashanksheela]

So what can I try to make it run?

[joshmoore]

Do you have another system that you can try on? Do you have another file system? What's the output of df? How did you install docker? What version?

[shashanksheela]

docker and docker-compose are at latest versions and using ubuntu 16.04
and df
Filesystem 1K-blocks Used Available Use% Mounted on
udev 8119620 0 8119620 0% /dev
tmpfs 1629240 9736 1619504 1% /run
/dev/sda2 944403884 84320416 812087412 10% /
tmpfs 8146188 34088 8112100 1% /dev/shm
tmpfs 5120 4 5116 1% /run/lock
tmpfs 8146188 0 8146188 0% /sys/fs/cgroup
/dev/sda1 523248 4756 518492 1% /boot/efi
tmpfs 1629240 60 1629180 1% /run/user/1000

[joshmoore]

docker and docker-compose are at latest versions and using ubuntu 16.04

Go ahead and paste the version numbers please.

One thing you might try is:

docker-compose up -d

and then

docker-compose down

i.e. don't use Ctrl-C, but I'm really grasping at straws until I can see the problem happen here.

[shashanksheela]

Docker version 18.03.1-ce
compose file format 3.6
docker-compose version 1.21.2

docker-compose down will work because it creates new container

[joshmoore]

docker-compose down will work because it creates new container

What's the harm in that? Isn't it only the volume that you care about?

[shashanksheela]

Actually what usually happens is our app will be running and we run docker-compose up parallely so only necessary apps will get restart

[joshmoore]

Ok. I'm not sure how you're using it, but my suggestion wasn't likely to be different than the Ctrl-C anyway.

Just to double check: you're not mounting myconfigdata in any other container, correct?

[shashanksheela]

yes i am not mounting in any other container

[joshmoore]

Unless you think you can show me more of your compose file, I'm not sure I have anything else to go on. You could try taking docker-compose out of the picture and test strictly with a docker image and see if it will work for you.

[shashanksheela]

docker image is working perfectly. Anyway, I will just think about alternatives. Thank you @joshmoore for your guidance.

[joshmoore]

You're very welcome, @shashanksheela. If you find a solution, please do let us know. (As a side note: if you happen to also be using kubernetes you could try https://github.com/openmicroscopy/apacheds-k8s-app)