Start a Conversation

Unsolved

This post is more than 5 years old

2755

September 19th, 2017 18:00

Scaleio scini.ko service issue

Hello,

I have configured a 3 node cluster on Ubuntu 16.04. The original plan was to have a converged cluster with all machines acting as both SDS and SDC. When installing, the Installation Manager correctly went through the process of installation with no errors reported even with the SDCs. However, the GUI displays 0 defined SDCs.

I have manually tried to add SDC with the following steps:

                        1. In new host (156.59.48.202), dpkg -i EMC-ScaleIO-sdc-1.32-2451.4.el6.x86_64.deb

2. In MDM, scli --add_sdc --sdc_ip 156.59.48.202

3. In new host (156.59.48.202), modify /bin/emc/scaleio/drv_cfg.txt, to specify MDM IP

4. /etc/init.d/scini restart


After attempting the final command of restarting the scini service, The error below is received: scini.ko restart error.png

I have used ScaleIO scini.ko kernel version mismatch as a reference guide as the issue is somewhat similar, although the output of the /etc/init.d/scini start command is slightly different. Assuming it is the same kernel version mismatch problem I am also unable to find the required correct kernel version file for scini.ko. I am currently running 4.10.0-35-generic and am unable to find that version within ftp://ftp.emc.com/Ubuntu/

My drv_cfg.txt file contains the following:

#Configuration file for the kernel module.

# One line stating the client GUID.

#ini_guid d9b8a46d-4abd-4ec9-abb2-7bd74f6b0ebe

# Additional line for MDM (Meta-data servers)

#mdm 10.20.30.40

ini_guid 43073f0f-8490-4085-bbbf-ba6d249a56d8

mdm 156.59.48.201,156.59.48.202

The MDM entries are correct as these are my primary MDM and secondary MDM.

The other file I have attempted to edit is /bin/emc/scaleio/scini_sync/driver_sync.conf with the following information:

repo_address = ftp://ftp.emc.com/

repo_user = scini

repo_password = scini

module_sigcheck = 0

After attempting those changes, and trying prior to the changes, the problem persists. Appreciate any advice pointing me in the right direction.. Thanks





306 Posts

September 21st, 2017 03:00

Hi,

Is there any particular reason you are using ScaleIO 1.32? It is very old and in 2 months will no longer be supported. I would suggest you to reinstall the cluster using the newest available 2.0.1.x code.

Thank you,

Pawel

4 Posts

September 21st, 2017 13:00

Hi Pawel,

My apologies, I had taken that step by step process from a previous article, whereas in my situation I am indeed using the newest version of ScaleIO.

Any further advice? Thanks

306 Posts

September 22nd, 2017 01:00

Hi,

I can't see you kernel version in the repository at all, so that's most likely the cause of why the SDC is failing to install the module. Let me ask around what's the roadmap for 4.10.x.x and I will get back to you.

Cheers,

Pawel

4 Posts

September 24th, 2017 11:00

Thanks Pawel, would downgrading the kernel version to one that is within the repository be a viable solution here?

306 Posts

September 25th, 2017 02:00

Hi,

Yes, I am pretty sure that SDC can't start becuase it can't fetch modules for this particular kernel version - if you can try one of the kernels you can find the drivers for in the FTP repo, that would be great.

Cheers,

Pawel

4 Posts

September 26th, 2017 18:00

It worked perfectly after downgrading the kernel version to one which was within the FTP repo.

Also noticed after doing this that the repo was actually updated today with the 4.0.10-35-generic kernel version which I was after.

306 Posts

September 27th, 2017 02:00

Just a bad (good?) luck then - I am not really sure how often the repo is being updated. Anyway, I am glad it worked for you!

Cheers,

Pawel

No Events found!

Top