Unsolved
This post is more than 5 years old
16 Posts
0
1536
September 1st, 2016 05:00
ScaleIO version 2.0.0.1 OS crash ( reboot ) bug at IM installation
I am trying to install ScaleIO_2.0.0.1 with Full IM mode on Redhat 7.2.
But I've not been able to complete the installation. At the install phase after the query and upload phase compleation, target OSs automatically reboot before its completion. It looks the reboot occurs after the MDM package's installed.
Is the reboot the right process during the IM installation? or any way to stop the reboot from happening?
My topology have 3 physical servers with RHEL7.2 and the each has MDM, SDS and SDC nodes.
So the reboot after MDM installation interrupts the following other node installations.
(Retray fail could work for the intall phase but the next configure phase fails after all)
I checked up the installation and user guides and found nothing noted for the reboot.
In addition, every ScaleIO installation demo video goes through without the reboot intervention during the IM installation.
Although the videos are not of version 2.0.0.1(mostly 1.32.xx).
I appreciate any help.
Taizo
pawelw1
306 Posts
0
September 1st, 2016 05:00
Hi,
It sounds like a bug we encountered with "drv_cfg" (which is part of the SDC installation) - can you please try to go for 2.0.0.2 or use the workaround provided in the KB article (installing 2.0.0.2 is more recommended):
https://support.emc.com/kb/486909
?
Many thanks,
Pawel
c0redump
68 Posts
0
September 1st, 2016 10:00
Hi Pawel,
We are currently evaluating ScaleIO, I found some bugs in the version v2.0.0.1 so I'm waiting the release of free version v2.0.0.2 to continue our evaluation session. On the download page I can only see the version v2.0.0.1 available, the v2.0.0.2 you are referring to is still unavailable.
Since you are talking about the version 2.0.0.2 I'd like to know if I can get that version somewhere.
Thanks in Advance,
Davide
hakuden
16 Posts
0
September 1st, 2016 17:00
Hi Pawel,
Thank you for your reply.
I tried the workaround on the version 2.0.0.1 written in the link you gave me.
Since there seems no free version 2.0.0.2 available.
https://support.emc.com/kb/486909
But it didn't solve the bug. OSs still keep crashing at SDC installation.
I tried the followings
1, added the following parameter to /etc/sysctl.conf
kernel.panic_on_oops=0
ran "sysctl -p" to load above setting.
2, rebooted after 1, and repeated 1 again
If there were no workaround for sure now, I would have to try an older version.
Many thanks,
Taizo
pawelw1
306 Posts
0
September 4th, 2016 13:00
OK, thank you for the information - to be honest I didn't know that 2.0.0.2 is not publicly available yet - can you please try again on Monday and see if you can download it?
Cheers,
Pawel
SanjeevMalhotra
138 Posts
0
September 4th, 2016 13:00
Hello Taiz
I have a few questions:-
1. Which all packages finish installing when the OS crashes?
2. Can you please run the command "rpm -qa | grep -i emc" and let us know which packages got installed on your node?
3. Does the install of SDC finish or does the OS panics/Crash even before the SDC install finishes?
4. Is the issue occurring only on this particular host or does it also occur on other hosts as well?
5. What is the model of the server/vendor you are using. Can you please verify that the hardware compatible with the Kernel version of your OS?
hakuden
16 Posts
0
September 5th, 2016 19:00
Hi Pawel
We talked to EMC, and got a FTP download link to the Version 2.0.0.2.
So we're now trying to install the 2.0.0.2 but got into a different problem.
Please check the post a teammate of mine put.
https://community.emc.com/message/950297#950297 <https://community.emc.com/message/950297#950297>
Hi SanjeevMalhotra
I will answer to your questions.
1, 2
Only MDM and SDC are installed (I ran a command "yum list installed | grep EMC").
On IM the status stops at MDMs completed, SDCs runnning due to the disconnection by the OS panic.
3
So it is fair to say that the OS panic occurs right after SDC installation.
4
It occurs on all the 3 hosts in the ScaleIO system. The 3 hosts have a SDC.
5
The server model is NEC Express5800 R120g-1M
To be more presices about the OS panic after I changed the sysctl.conf(kernel.panic_on_oops=0).
The srevers stoped rebooting automatically but they just froze and never to come back even after HW rebooting. So We need to do OS re-installations.
Many thanks
Taizo
osaddict
110 Posts
0
September 6th, 2016 09:00
The F&F download link has been updated with the 2.0.0.2 version. The website will be updated shortly to reflect that change on the website.
KreizO
2 Posts
0
September 6th, 2016 23:00
Hi,
I had the same issue with 2.0.0.1 and Centos 7.2. Fixed only with newer kernel from ML repo.
smatsuda
14 Posts
1
September 7th, 2016 17:00
Hi
We now use the 2.0.0.2 version and the installation is successful.
Thank you for the quick response.
Cheers
Seiji (a team mate of Taizo)
pawelw1
306 Posts
0
September 8th, 2016 01:00
Happy it worked, thank you for the info!