Unsolved
This post is more than 5 years old
10 Posts
0
25496
September 17th, 2015 11:00
Detect network hardware causes iDRAC to disconnect
I have a Dell Poweredge R610 server which is connected to a 10.0.0.0/29 network. This server has iDRAC enterprise and the NIC selection set to shared with static IP address 10.0.0.6.
I'm trying to install Ubuntu on this server and during the "Detect network hardware" step iDRAC's IP (10.0.0.6) becomes inaccessible and I need to use KVM-IP to proceed further. Until the server is rebooted the IP stays inaccessible. Why is it so?
iDRAC firmware revision is 1.92.05
No Events found!


DELL-Josh Cr
Moderator
•
9.6K Posts
•
42.5K Points
0
September 17th, 2015 14:00
Hi,
It may have to do with it resetting the NICs during the discovery phase. If you use the dedicated port does it do the same thing? You may want to update the firmware as well. http://www.dell.com/support/home/us/en/04/Drivers/DriversDetails?driverId=0F12K&fileId=3470464256&osCode=RH60&productCode=poweredge-r610&languageCode=EN&categoryId=ES
jesin
10 Posts
0
September 17th, 2015 15:00
Hi Josh,
Thanks for the quick response.
I updated iDRAC to 1.99 but the problem still persists.
I haven't tried this but I wish to use the shared option.
Should I update the firmware of the network device too?
Thanks,
Jesin
DELL-Josh Cr
Moderator
•
9.6K Posts
•
42.5K Points
0
September 17th, 2015 17:00
It wouldn’t hurt to try it. When the idrac stops responding how long have you waited? The idrac may need to reboot which can take a few minutes.
jesin
10 Posts
0
September 17th, 2015 17:00
I didn't notice how long I waited but it must be atleast 10mins.
Also when the server boots via PXE, it gets the same IP which was assigned statically to iDRAC. And this causes iDRAC to disconnect.
DELL-Josh Cr
Moderator
•
9.6K Posts
•
42.5K Points
0
September 17th, 2015 18:00
10 minutes should be enough for it to reboot. The duplicate IP addresses could be an issue.
jesin
10 Posts
0
September 18th, 2015 07:00
I booted from a CentOS live image (via virtual image) and updated Broadcom's firmware but the issue was still present. So i decided to update the BIOS.
I booted from the same CentOS live image (via virtual media), downloaded the BIOS BIN file, executed it, it completed successfully and asked to reboot. I hit 'y'
Now CentOS is stuck here - http://i.imgur.com/YxCOceU.png
But the system hasn't frozen and I can issue a Ctrl-Atl-Del to it. And when I do this the "init: plymouth-shutdown..." line appears and it stops at "init: rc main process killed..."
What do I do now? iDRAC isn't accessible and I'm seeing this output screen via KVM.
DELL-Josh Cr
Moderator
•
9.6K Posts
•
42.5K Points
0
September 18th, 2015 10:00
will it let you reboot with ctrl-alt-del from that screen?
DELL-Josh Cr
Moderator
•
9.6K Posts
•
42.5K Points
0
September 18th, 2015 13:00
Can you change the IP so that there isn't an IP conflict with the iDRAC and the PXE? That could be causing things to not communicate properly.
jesin
10 Posts
0
September 18th, 2015 13:00
I had to power cycle the server, luckily nothing bad happened but the BIOS was back to the older version.
Then I mounted the ISO via KVM over IP and was able to successfully update the BIOS. But iDRAC still becomes inaccessible during the "Detect network hardware" step.
Also the iDRAC webpage > Firmware Inventory shows the old Broadcom firmware version. I thought it wasn't updated and ran the .bin file again. But the "Running validation...." step shows the latest firmware version.
So why is iDRAC still reporting the older version?
jesin
10 Posts
0
September 20th, 2015 09:00
I even disabled DHCP for this network but it still didn't work. I came to the conclusion that a dedicated link is required as a normal NIC is under the control of its OS.
Example, executing "ifconfig em1 down" from within the OS shuts down the port including the shared iDRAC.
So got another link from the switch to the dedicated iDRAC port and all is well.
Thanks a lot for the support Josh!