Not able To add start two instances of version 3.1.0 on same machine

[alok-agarwal]

Hello @carlosforero ,

Even I am facing the same issue. Multiple instances of version 3.x.x wont run simultaneously.

Although, if we try to run one instance of version 2.x.x and one of 3.x.x, it works fine.

Hope that helps you in some way. Looking forward to resolution of the issue.

Thank you!

[carlosforero]

Thanks @alok-agarwal. I'll check it out today.

[alok-agarwal]

Thank you @carlosforero for this really handy instance manager. Appreciate it!

Some more insights, I tested on Centos and Ubuntu but the issue is replicated on both.

Today I upgraded ineo to the latest version but the issue persists. The difference i could see is that correct port numbers and status is reflected when we do ineo instances or ineo status.

Many thanks!

[carlosforero]

@alok-agarwal

I tried to reproduce the problem, but has not been possible until now.

I'm testing on Mac and Ubuntu 16.04.

$ ineo status

  status 'a310'
  Neo4j is running at pid 3333

  status 'b310'
  Neo4j is running at pid 3396

  status 'a331'
  Neo4j is running at pid 3463

$ ineo instances

  > instance 'a310'
    VERSION: 3.1.0
    EDITION: community
    PATH:    /home/ubuntu/.ineo/instances/a310
    PORT:    9495
    HTTPS:   9494
    BOLT:    9496

  > instance 'b310'
    VERSION: 3.1.0
    EDITION: community
    PATH:    /home/ubuntu/.ineo/instances/b310
    PORT:    8484
    HTTPS:   8485
    BOLT:    8486

  > instance 'a331'
    VERSION: 3.3.1
    EDITION: enterprise
    PATH:    /home/ubuntu/.ineo/instances/a331
    PORT:    7474
    HTTPS:   7475
    BOLT:    7476

I can also open the Neo4j web manager for each instance.

So, please can you paste your ineo instances output, ineo status output, and the last lines of /home/[your_username]/.ineo/instances/[your_instance_failing]/logs/neo4j.log ?

[alok-agarwal]

@carlosforero

The output for ineo instances and ineo status is same as yours.

And the neo4j.log only has an entry only for the start of the instance. No error as such in there. So, all the instances are running.

But, when you try to access these three from the browser it will only work for the first instance and not for the other two instances. Also, if you try to connect to these instances from some scripting language like php, it does not work.

[carlosforero]

I can reproduce exactly the same problem reducing the host memory to 720MB. When I set 1024MB then everything works ok. This host VM was installed just to use ineo.

The free output using 720MB (before start neo4j instances) is:

              total        used        free      shared  buff/cache   available
Mem:           700M         50M        169M        2.3M        479M        501M
Swap:            0B          0B          0B

The free output using 1024MB (before start neo4j instances) is:

              total        used        free      shared  buff/cache   available
Mem:           992M         40M        805M        3.1M        146M        802M
Swap:            0B          0B          0B

So, I think the problem is related to Neo4j instead of Ineo. If you try to start the instances without ineo, you possibly get the same problem.

Let me know if those findings apply in your case.

[alok-agarwal]

Hi @carlosforero ,

The machine memory on which I have these instances is way more than 1024MB.

The free -m output (before start neo4j instances) is:

              total        used        free      shared  buff/cache   available
Mem:          10741        1452        6875          85        2413        8852
Swap:         10996           0       10996

The free -m output (after start neo4j instances) is:

              total        used        free      shared  buff/cache   available
Mem:          10741        3268        5056          86        2416        7035
Swap:         10996           0       10996

Is there any setting I am missing out on? Tried setting the dbms.memory.heap.initial_size to 1024m, but that did not help. Looking forward to your help in getting these started.