-=DoW=- Dsotm
June 9th, 2003, 09:43
This from SCI VP over at AA forum:
Posted: 2003-06-09 01:53:25
--------------------------------------------------------------------------------
Hi everyone! I'm sitting here at the new Data Operations center for the SCI systems. I wanted to let you all know that we have been working on getting the clusters up and going for approximately 36 hours now without a pause. I saw some things here in the forums and wanted to address those quickly.
1) Will we be prorated on the leased servers?
Yes, all of the server admins had an email sent that told them that we are not only prorating the time that the servers are down, but actually giving DOUBLE credit. So if the total time down is 48 hours, the leased servers will be credited for 96.
2) Why did they choose a weekend to do the work?
We really weren't given a choice on times. When you move from one data center to another, you have to arrange the schedule so that BOTH providers are ready for the change.
3) Why does the ETA keep changing?
As we are going through this move, we are also upgrading much of the network equipment such as switches, routers, etc. New equipment sometimes fails and needs to be replaced on the fly. We had one VERY major piece of equipment that the vendor actually had to hand carry to us on a plane! We are working around the clock to get your systems back up and running.
4) Why didn't you leave a part of the network up and move it a piece at a time.
The network that we run is NOT like a normal network. This is a HUGE cluster of servers that act as if they were one machine. To give an example, imagine you were moving your home machine. Now on day 1, you move the monitor and a single hard drive... Day 2 you move the power supply and the RAM. You get the picture? The cluster had to be moved as a whole or we would just have a longer down time.
The end result is that we feel that you will be EXTREMELY pleased with the result of this move and upgrade.
We appreciate your patience and look forward to your comments after the cluster comes back online.
(1VB)compforce
Senior Developer, SCI
[ This Message was edited by: (1VB)compforce on 2003-06-09 01:55:32 ]
Full thread here: http://americasarmy.com/forum/viewtopic.php?topic=68334&forum=15&16
Posted: 2003-06-09 01:53:25
--------------------------------------------------------------------------------
Hi everyone! I'm sitting here at the new Data Operations center for the SCI systems. I wanted to let you all know that we have been working on getting the clusters up and going for approximately 36 hours now without a pause. I saw some things here in the forums and wanted to address those quickly.
1) Will we be prorated on the leased servers?
Yes, all of the server admins had an email sent that told them that we are not only prorating the time that the servers are down, but actually giving DOUBLE credit. So if the total time down is 48 hours, the leased servers will be credited for 96.
2) Why did they choose a weekend to do the work?
We really weren't given a choice on times. When you move from one data center to another, you have to arrange the schedule so that BOTH providers are ready for the change.
3) Why does the ETA keep changing?
As we are going through this move, we are also upgrading much of the network equipment such as switches, routers, etc. New equipment sometimes fails and needs to be replaced on the fly. We had one VERY major piece of equipment that the vendor actually had to hand carry to us on a plane! We are working around the clock to get your systems back up and running.
4) Why didn't you leave a part of the network up and move it a piece at a time.
The network that we run is NOT like a normal network. This is a HUGE cluster of servers that act as if they were one machine. To give an example, imagine you were moving your home machine. Now on day 1, you move the monitor and a single hard drive... Day 2 you move the power supply and the RAM. You get the picture? The cluster had to be moved as a whole or we would just have a longer down time.
The end result is that we feel that you will be EXTREMELY pleased with the result of this move and upgrade.
We appreciate your patience and look forward to your comments after the cluster comes back online.
(1VB)compforce
Senior Developer, SCI
[ This Message was edited by: (1VB)compforce on 2003-06-09 01:55:32 ]
Full thread here: http://americasarmy.com/forum/viewtopic.php?topic=68334&forum=15&16