cornelius785
join:2006-10-26
Worcester, MA
| no backup?
seems awfully strange that there are no backups of the email. either charter is a really crappy company and pretty stupid or the writer is putting some highly opinionated and speculatory BS in the news post, not sure which though.
i'd like to hear from charter on what happened. if it was just maintenance, i bet someone lost their job. there are too many unknowns right now. does charter have multiple duplicate data centers? they are using RAID, right? and more. |
|
Pictor Guy
join:2004-06-21
Sammamish, WA
| RAID is not backup. RAID only provides fault tolerance in the event of a drive failure and perhaps some performance gains depending on the setup.
Backups aren't fool proof either. If there is no testing of recovery you could be backing up thinking you can restore but when the excrement hits the fan then there could be the realization that the backups failed too. Either way this sort of thing happens but is usually a result of poor planning, poor management, and poor disaster recovery plans. I suspect their Risk Management team is doing some spin control for the public and insurers, not to mention how they may have to deal with legal issues as a result of ongoing warrants or investigations. |
|
raw
War Eagle
Premium
join:2001-01-17
Pasadena, MD
clubs:
| reply to cornelius785
said by cornelius785  :seems awfully strange that there are no backups of the email. either charter is a really crappy company and pretty stupid or the writer is putting some highly opinionated and speculatory BS in the news post, not sure which though. As a dissatisfied customer, I'd be more than willing to put my money on the former. |
|
RARPSL
join:1999-12-08
Suffern, NY
| reply to Pictor Guy
said by Pictor Guy :RAID is not backup. RAID only provides fault tolerance in the event of a drive failure and perhaps some performance gains depending on the setup. Some RAID configurations can be set as backups. The way it works is that the configuration of duplicates is such that the data is recorded 3 times (on 3 separate drives in the RAID complex). You disconnect the drives that comprise one full image from the RAID Complex and access it as separate image for your backup run [ie: It is a snapshot as-of the time of disconnect]). Once the backup run completes, you reconnect the drives to the RAID complex and it will update them as if you just HOT SWAPed an empty drive. The problem with doing backups is that you need to freeze the drive during the backup so you do not get an in-fight change as you do the backup. |
|
uncleFester
@rr.com
| said by RARPSL :Some RAID configurations can be set as backups. The way it works is that the configuration of duplicates is such that the data is recorded 3 times (on 3 separate drives in the RAID complex). You disconnect the drives that comprise one full image from the RAID Complex and access it as separate image for your backup run [ie: It is a snapshot as-of the time of disconnect]). that's just one method (and these days, one of the more costly/wasteful methods). plus, you only have one fallback position so what interval do you take it? hourly? daily? the RPO/RTO demands will dictate this period, along with the capabilities of the technology used.
honest-to-god snapshotting (say, netapp's snapshot or emc timefinder/snap) can do this kind of thing better.. but it's more technical, depends on a lot more pieces working correctly and has its own impacts on response/service levels.
if i had to bet money, charter does back this up. the failure simply happened in the window between backups, the loss was irrevocable. think what happens when your backups run at 8PM and your disk fails 7:59PM. life sucks. any company faces this risk. it simply depends on how much value is placed on the data and the risk if data is lost. being email, the value is low and the risk is negligible. |
|
