New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Most of the time a FSCK is not needed with hardware raid controllers, they will detect the drive has been replaced and rebuild the array transparently to the OS (well, things will crawl for a while, but in some cases with plenty IOPS to spare, it may go unnoticed as rebuilding array can have a low priority and, especially in raid10, it will also be quick).
The fact a FSCK was needed is probably the consequence of the "multiple" word which speaks of a high risk of data loss if not already some lost, so rebuilding the array with 0 redundancy under load as well as not doing a check were out of the question options, but they can be considered in normal cases (when only one drive failed).
Actually 13 hours on 40TB of raw storage is nothing short of rapid, although why you have that much storage on a single node is beyond me
It's probably a SAN. Or some 4U monstrosity
Yeah.. but that's an E3 node I thought..
In cases where you do a back up from one server to another on a different network, would you switch the client domain to the backup server if the primary server died?
We stopped using E3 for OVZ a while back. NL4 is 2U HP DL380p gen8 with dual e5-2650
http://iniz.com/blog/amsterdam-infrastructure.html
This would be high availability cloud with SAN and is no way possible/profitable within LEB pricing
Cool. One of the reasons why I love your service. It was awesome that you added an extra ipv4 to existing clients without any fanfare or ado.
That said, I'd never rely on a host to take backups for me. I have crons setup to rsync tars with a Backupsy. Hopefully both of them won't fail at the same time. They're both US though different DCs, but you never know with natural calamities.
It seems tun/tap is not working correctly yet?
Please submit ticket if your still having issues, try disabling and re-enabling from CP as well.
Okay... can't i set up a cloud flare to switch between A records if one is found offline
That is what every sane person should do. We do take some backups but we consider it is better to give the storage space and the customer to setup them as they need, when they need. It distributes the load evenly and it eliminates the issues with wrongly timed back-ups for almost everyone. Nobody can know better than the user when to take backups and what to include in them, taking the backups at provider level is a big waste of space and IOPS/traffic/CPU for almost nothing as the timing of the backup will not make anyone happy, nor the intervals.
@Iniz is one of my top 3 providers, i have a few hundred days uptime