up vote 1 down vote favorite
share [g+] share [fb]

I've been preparing a new Linux server in the office to be used as a VM host using Proxmox. Now that the basics are done we're moving it to our ISPs server room.

I'd like to make sure I haven't forgotten anything before we move the server. Here are things I've thought of:

  • configure network to use correct static IPs
  • update /etc/hosts, /etc/hostname, run hostname
  • verify NAT config from ext. network
  • update HW service info
  • verify monitoring/error reporting (HW+SW)
  • label drives+cables
  • backup system image (just in case)

Is there anything missing from this checklist? It's on a pretty general level to make this question maybe helpful for others too.

link|improve this question
86% accept rate
feedback

3 Answers

up vote 2 down vote accepted

You've said to update the IP address, but not mentioned other things that hang off it (default gateway is very often overlooked, accessible DNS servers are another). My apologies if you considered that implicit in changing the IP.

I also understand from others' answers that this is a physical server you're moving (I thought from the question it was moving to be a VM image, but I'm just an idiot!). If it's a physical box, something that's often overlooked is establishing the procedure for getting physical access to it in the datacentre (who do I call? do attendees need to be on a pre-written list? do we need any phyiscal doco (photo ID, and if so, what sort; employer ID?)). The right time to find these things out is not 4am, when the motherboard has just failed.

In respect of that latter, I've been bitten enough with these procedures that I always now test them out. Call the number at 3am, and check that you and Fred the PFY will be able to get access to the box in 45 minutes. When it turns out that noone answers that number in the middle of the night, or that you can only get access in business hours, or that Fred's not on the night visitors list, you can escalate that before it's of critical urgency.

link|improve this answer
oooh 8192 rep get a screenshot of that :) – Iain♦ Aug 11 '11 at 10:38
That was implied by the first item on the list, but you're right and I should've mentioned these separately. Since the ISP stays the same, the DNS servers stay the same (verified this) but the gateway changes. – onik Aug 11 '11 at 10:40
Iain, you'll be at 16384 shortly, don't forget your screenshot when the time comes! – MadHatter Aug 11 '11 at 11:05
Physical access is a VERY good point, would +2 if possible. Have to verify this ASAP. – onik Aug 11 '11 at 11:31
feedback
up vote 1 down vote

Check that the power in that rack is compatible with your hardware, there is enough of it and the power cables are compatible. Is there sufficient space in the rack.

link|improve this answer
I've prepared for both C13/C14 cables and IEC cables, power is supplied from a sufficient UPS and there's 4U of rack space reserved (only 2U needed). I'm going to check the rack and install the rails before actually moving the server. – onik Aug 11 '11 at 10:58
feedback
up vote 1 down vote

Check BIOS settings (power on after power failure), and remove any CD's from CD-ROM ;-)

link|improve this answer
The server powers to the previous state after power failures, and boot priority takes care of "Boot from CD" problems. Of course one should not store CDs in the CD drive ;) – onik Aug 11 '11 at 11:28
feedback

Your Answer

 
or
required, but never shown
discard

Not the answer you're looking for? Browse other questions tagged or ask your own question.