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
RAID, fingers crossed.
When shit hits the fan, cry on LET and put all blame on my provider.
Do it more frequently.
Also people seem to use incremental backups. I don't.
KS2 with 2TB HDD and a cluster of raspberry Pis with I think a total of 4 1TB HDDs. Might be 3 1T and 1 750G
No strategy. I like to live on the edge.
BackupPC at 2 locations
No backups, as I've only got a cluster with idle servers - with the same on each one of them - nothing as of a ssh config and the files created by the OS.
run crying around the block.
or cry running, its ok too.
i have 4TB nas at home, 3TB backup in a VPS, 1TB dropbox, 1TB google drive
so i mostly use VPS, and drive.
some most important at NAS, DRIVE and DROP.
Rsync and Attic depending on what I'm backing up.
Kimsufi boxes.
RAID and Pray to God that it never screws me over.
RAID is not backup.
For files I occasionally rsync to other servers. For database I use MySQL replication.
Personally; a RAID NAS box at home with incremental live sync offsite to multiple boxes around the world. Version control is done at a file level.
It's backwards to talk about backups. What scenarios are you trying to address? You should be talking about restore strategies. What are you planning to restore with a sub-daily turnaround?
That's why Apple's Time Machine is so handy. Behind the scenes it's just doing an incremental rsync. The restore interface, though, is what makes it useful to have access to hourly backups.
Myself, lacking a more useful front end, I just keep daily local backups of important directories/files. An example:
And they get replicated remotely on a weekly (or sometimes monthly) basis. Increase the frequency if you really have something more important, but don't kid yourself into thinking that just having a recent duplicate of the data means that your job is done.
Rsync
Rrsync is great if you don't want to make chroot accounts. Works fine on debian. It restricts local access to the rsync server for rsync only.
The rsync server's ~/.ssh/authorized_keys looks like this:
Example:
root@server:~# ssh namefromsshconfigfile PTY allocation request failed on channel 0 perl: warning: Setting locale failed. perl: warning: Please check that your locale settings: LANGUAGE = (unset), LC_ALL = (unset), LANG = "en_US.UTF-8" are supported and installed on your system. perl: warning: Falling back to the standard locale ("C"). /bin/rrsync: Not invoked via sshd Use 'command="/bin/rrsync [-ro] SUBDIR"' in front of lines in /home/username/.ssh/authorized_keys Connection to IP-address closed.Rsync with rrsync:
opening connection using: ssh namefromsshconfigfile rsync --server -vvlograndomchars . /rsync/remotehost/backup/directory/ perl: warning: Setting locale failed. perl: warning: Please check that your locale settings: LANGUAGE = (unset), LC_ALL = (unset), LANG = "en_US.UTF-8" are supported and installed on your system. perl: warning: Falling back to the standard locale ("C"). sending incremental file list delta-transmission enabled 20151210file1.tar.gz.gpg 20151210file2.tar.gz.gpg 20151210file3.tar.gz.gpg 20151210file4.tar.gz.gpg total: matches=0 hash_hits=123 false_alarms=0 data=123456 sent 206889 bytes received 3766 bytes 28087.33 bytes/sec total size is 206580 speedup is 0.98I just set up 3 backup box using rsnapshot, it use hard link to do increamental backup, so if I want to restore, just copy hourly.0/files
RAID is good. But it's not the backup
My backup plan is lacking at the moment but I plan on improving it over the next few months now that I have more hard drives available.
What's used for real time sync?
Closest I can come is using 3rd party stuff(dropbox, owncloud and such) or rsync every minute.
On my Mac is Arq and ObjSpace, for servers Duply/Duplicity and ObjSpace.
Playing with Hashbackup and ObjSpace at the moment.
Why is RAID not a backup?
Quote: "...If you accidentally overwrite your PhD thesis with garbage, redundancy ensures that you have multiple copies of garbage, in case one gets bad. A backup ensures that you can restore your PhD thesis..."
I use carbon copy cloner for backing up my mac via a full incremental disk image.
Synology Cloud Station
Everything goes to my Synology DS215j (2x 3TB RAID0) twice a day. Synology NAS syncs with Google Drive.
I also rsync some stuff to my homeserver (2x 120gb ssd RAID1, 2x 3tb hdd RAID0)
Plan on getting a Qnap TS-431 in the future to have some more space.
I like how convenient Qnap and Synology are. I could stuff 8 more HDDs into my homeserver but I like it more to have a NAS that handles everything for me.
R1soft and pray to god it works when you need it
Home boxes and stuff hosted on externally push data to a couple of cheap dedi/vps accounts daily via rsync. Larger dedi/vps accounts pickup copies of those backups daily and maintain 8 of each daily, weekly and monthly snapshots. Nothing massive in that arrangement: things like TV/film/music files can for the most part be retranscoded/redownloaded from elsewhere easily, I only bother backing up difficult to replace stuff (work files, source repos, my own photos & videos, machine/app configuration, anything that lives in places like /home, and so forth). A few special case items like full copies of VMs are made manually after major updates. That backup machines have their configuration and other stuff backed up the other way around, and everything has local snapshots as well as the ones in the backups.
Pushing to a VPS that the backup hosts then pull from means that the source and backup servers don't need to be able to authenticate with each other, only the middle-man services. This means that if my main machines are hacked it isn't so easy for the bot/hacker/kiddie to break the backups as well (similarly breaking into the backups or intermediates doesn't give you access to the source machines). I just need to keep the keys safely locked (otherwise that effort decoupling the sources and backup copies is wasted).
For really vital things (finances, keystore, crown-and-jewels source code, contact details) I have copies on USB sticks as truly offline backups, at least one here, one secreted in a drawer at work, one that a friend looks after. These get cycled every week-ish. Said friend's machines backup to my machines too.
I have at least four copies of just about everything, not counting all the snapshots. Possibly over the top, but it makes my data feel safe-ish. It would take quite a disaster (or a catastrophic cock-up on my part) to completely lose something I care about.
Oh, and some stuff on OneDrive, mainly because my phone does that automatically, partly for sharing certain documents between the main desktop, table, and sometimes phone.
And just about everything lives on either RAID1, 10, or 5, depending on the expected access patterns and the kit it is on, rather than individual drives.
My backup strategy:
Pfft.. who needs backup for running IO benchmarks all day?
But seriously I just have a script to zip the entire / var/www/ folder and mysql dumps, then upload it to Dropbox.