Bladewatch.com

Datacenter planning problems (part1)

The key issues I’ve dealt with, (in partnership with datacenter planning, unix and windows teams) have been:

* Physical/virtual records inaccurate – box migrated to vmware, but physical server redeployed – in effective you now have two servers linked to one asset

* Server has been rebuilt and allocated a new hostname – this might not be reflected in the different records – to purchasing it’s a box, to windows team it’s server 123, and to datacenter planning it might be one of the many DL380′s or just asset 7.

* Datacenter relocations – your records might say it’s in bcp site 1 – when in fact you’ve moved datacenters as part of a bcp migration project

* Internal systems might disagree on ‘production/development’ status – status is key, taking a production box out of service could cause disruption to service, loss of business or damage to the company image – make sure everything’s verified by the service delivery teams

* Operating system/hardware might be incorrect, your internal records might show it’s windows nt or solaris 7 when the box has since been rebuilt with a new operating system, or the hardware supplemented with another server, for example a 6500 being swapped with a DL360. The asset may be linked to two hostnames now.

* Remedy/helpdesk/change process might not be compatible with real life, if the process takes too long or doesn’t ask the right questions, you’ll never log the information you want.  Server decommission, note asset tag, server type, operating system etc.

These are the main ones of the top of my head. Any issues/comments on the issue outlined would be welcomed.

No related posts.

Related posts brought to you by Yet Another Related Posts Plugin.