Re: ReactOS

Publié par Robert le 21/12/2006 19:39:42

Effectivement, le site dédié présente un problème, voir ci-dessous :
Failure Note
ReactOS Project
Latest News: Server Copy Ready, all data Recovered, HDDs on the way to the Hostingsite again
19.12.1006 IMPORTANT update: After 3 Times crashing of the Copy machine (1HE system) we switched to a normal PC. after 22 hours of "copy and dd" - but the copy process is still running ....
we have this status: - i do not think i have to comment this
Time Line and Status: (what we know so far)
Thursday 00.05 Raid Controller on HostSystem where Reactos Backend is running decided to do a Raid Rebuild.... (maybe triggerd by a dead hdd..unknown - Logfile is unclear about this)
Thursday 03.15 Raid Controller on Hostsystem Crashed the System, as it took Raidset offline
Thursday 09.00 Technicans investigated local to find the problem of the "crash" -- was unknown Why it crashed .. and came with the cool news "on reboot, the server writes: No operating system found..."
Thursday 15.00 It was clear, that the Raid controller inside the Server failed, and they "forgot" all Raid partitions. (reason unknown)
Background: Reactos is splitted into a frontend and Backup system, Backup System resides inside a Vmware Session. The problem is, that the Vmware Session (disk) is on this failed Raid Controller. (also the daily backups) -- only an old 1 Month backup is avaliabe but this is indeeed VERY old.
Atm we are dumping down the Data from the Raid Disks disk with dd to a Transfer disk manually.
The 3'rd.. 4'th .. attempt was already VERY successful, the filesystem mounted almost without errors, but it was still not 100%.
We are Sorry, that this takes long, but did you ever dd a 1TB system ? to a new hdd ?
Really, I have seen EVERYTHING in my life ...
But we have never seen a Raidcontroller, who kills the Raidset so brutal, that not even a new Raidcontroller sees the Raid partition. (looks like a Software glitch in the Raidcontroller)
Future
We work hard, that over the Weekend the System comes up again. (all necessary Parts needed are bought already)
The chance that all this happened everything without loosing any data is a clear 99%.
Even, if we could, we could give more details, Like: Type of Raidcontroller, and so. But i do not want to Release this public, as "Bugs" can happen everywhere, and even Raidcontroller Programmers are People who can do mistakes. (maybe it was a failure of the RAM on the Raidcontroller...)
But i can garantee you here, that the Hardware is not a kind of <10k$ system ....
So keep cool, stay tuned, and do not panic, and do some X-Mas Party
(Thats indeed one of the problems we have, as all people needed to work on this are logically "on holiday", "drunk", ... --- > as always .. - and logically we are hit by Murphy's Laws (Anything that can go wrong, will go wrong.)...)