|
Yeah I'd just copy poo poo off to a different server and take it from there...
|
#
?
Nov 17, 2016 21:20
|
|
|
|
| # ? Apr 18, 2024 03:24 |
|
|
unknown posted:Haha, 'inherited' another ancient box. Here's a prime example of why raid is good, and why you should have backups. I'd call that an example of why RAID is bad actually.
|
|
#
?
Nov 17, 2016 22:15
|
|
|
redeyes posted:I'd call that an example of why RAID is bad actually. Really? The redundancy actually worked in this case. Server is still usable - I'll totally admit that all the drives need to be replaced, as if this many have failed, then the rest are ready for the trash heap in the sky/ground. Data is safely removed from the server in backups and running clean on a new box. All done in a non-emergency timeline. Can't see how this is bad...
|
|
#
?
Nov 17, 2016 22:38
|
|
|
Speaking of Raid it's probably a waste of my time to bother with Raid 6 if I'm going to just have four 6 (or 8) TB drives with no extra space for hard-drives in the case. Might as well just use Raid 1 and two volumes in that case right? No real benefit to go Raid 6/SHR-2 other than one volume in this case. I bought one of those Xeon Dell Poweredge T20s for $250 and I'm going to use it as an Xpenology box.
|
|
#
?
Nov 17, 2016 22:40
|
|
|
havenwaters posted:Speaking of Raid it's probably a waste of my time to bother with Raid 6 if I'm going to just have four 6 (or 8) TB drives with no extra space for hard-drives in the case. Might as well just use Raid 1 and two volumes in that case right? No real benefit to go Raid 6/SHR-2 other than one volume in this case. The big difference is that with SHR-2/RAID-6 you can have any 2 drives fail with no impact to data availability. If you are using 2 mirrored volumes and the two drives that fail happen to be from the same mirrored set you've now lost access to that data.
|
|
#
?
Nov 17, 2016 23:06
|
|
|
havenwaters posted:Speaking of Raid it's probably a waste of my time to bother with Raid 6 if I'm going to just have four 6 (or 8) TB drives with no extra space for hard-drives in the case. Might as well just use Raid 1 and two volumes in that case right? No real benefit to go Raid 6/SHR-2 other than one volume in this case. If you do the math out and Monte-Carlo simulate it out, raid 6 is something like 3x as fault tolerant as raid 1+0 for an equal number of drives. On a raid 6, any 2 disks can die horribly without issue, but on a RAID 1 pool if two disks in the same pool go poof, you're hosed.
|
|
#
?
Nov 17, 2016 23:08
|
|
|
unknown posted:Really? The redundancy actually worked in this case. Server is still usable - I'll totally admit that all the drives need to be replaced, as if this many have failed, then the rest are ready for the trash heap in the sky/ground. The worst part of raid is the lazy fucks who don't do anything about drive failures asap.
|
|
#
?
Nov 17, 2016 23:11
|
|
|
Krailor posted:The big difference is that with SHR-2/RAID-6 you can have any 2 drives fail with no impact to data availability. If you are using 2 mirrored volumes and the two drives that fail happen to be from the same mirrored set you've now lost access to that data. Ah. Yeah didn't think about it that way, SHR-2 while still backing up the NAS it is. Was planning to rotate a couple 8tb Externals and also back up the irreplaceable stuff to the cloud (photos & documents mainly). Still need to figure out how I want to backup stuff from the various computers to the NAS though. I'd almost prefer something like windows 8.1/10 file history but it's a pain to go back 500 revisions and look for something if you don't remember what you named it. edit: Mainly since I'd like the files to be easily accessible on the backup usb external if I have to take it to a computer offsite (say a laptop) and not have everything compressed into an archive.
|
|
#
?
Nov 17, 2016 23:12
|
|
|
I work at a school and for a long time we hadn't had any real backups - used to use portable hard disks but they failed and were not replaced, weren't big enough to hold everything required anyway. I wanted to do server backups for some VMs as well as just backing up user data. We got a new server with a tape drive (LTO6) so in theory I should be able to start getting things backed up again. As I said, we're a school, we don't have much money and in any case I don't have authority to buy anything so I've been looking for what's free. The server has a number of VMs and so I've started looking at Veeam backup & replication free version to start out with. It apparently works great for zipping the VMs although I haven't been in a position yet to attempt a VM or file restore. Going to try that once I get the old server wiped and ready to reuse for testing. The part I'm having trouble with is using the tape drive. I've not used one before so I'm not sure if my issues are due to my dumb mistakes or if there are problems with the drive. I noticed the first issue after I tried copying the VM backups to the tape drive. The drive would go "offline" after apparently finishing the job, Veeam wouldn't recognise when I swapped tapes etc although the unit itself still had power and could eject tapes fine. It would come back online and be usable after restarting the server and this is obviously very far from ideal for a VM host to be needing daily restarts. At first I thought maybe I had to "close" the tape in Veeam somehow before ejecting it but couldn't really find any info on that. I had some success with "exporting" the tape before pressing eject but I think it might have been coincidental timing that the drive did not go offline those times. It did seem to start with that the drive would go offline every time I ejected the tape but really difficult to pinpoint if it exactly lined up with that or was just happening at some point near to that in time. I tried doing driver updates and even firmware updates but it doesn't seem to be helping - going to older firmware the drive seemed to stop going offline but couldn't read the tapes and also didn't want to eject the tapes. Going to a newer version the tapes became readable again but the drive started going offline again. Can't see any errors with the drive in the IDTD tape diagnostic software although this software can't see the drive either once it disappears from Veeam. VM host is a Dell T630 running Windows server 2012r2. A bit of googling only finds a scattering of posts about tape drives that go offline and I haven't found anything that really matches my situation so could it just be a dodgy tape drive? I don't want to put a ticket in until I'm reasonably sure that I'm not just screwing up somewhere in the process. Anyone seen anything like this? Anyone got any tips for someone starting out with no idea and trying to use Veeam and LTO6? Edit: IDTD? ITDT? I can't remember what it's called. Stoca Zola fucked around with this message at 04:50 on Dec 9, 2016 |
|
#
?
Dec 9, 2016 04:46
|
|
|
Tape drives going offline usually indicates a hardware problem or a fundamental incompatibility between the tape drive and the system it's attached to. In my experience, it's rarely the backup software at fault. Since the VM host is a windows box, check your system events. Might also contact the hardware vendor for the tape drive and see what they think. Edit: Especially if multiple programs can't see the drive once it goes offline. That strongly suggests there's something wrong with the tape drive itself. ConfusedUs fucked around with this message at 05:10 on Dec 9, 2016 |
|
#
?
Dec 9, 2016 05:07
|
|
|
Bingo! Event log reveals that the controller that the tape drive plugs in to has firmware issues and is erroring out and restarting, the timing of this coincides with times I lost contact with the tape drive. So I'll start there. Thanks! Edited to add: I ran a tape job over night, with the new controller firmware it got to the end without going offline and was even able to eject the tape at the end of the job which had never worked before. So the tape drive was never at fault most likely. Stoca Zola fucked around with this message at 03:58 on Dec 14, 2016 |
|
#
?
Dec 12, 2016 07:08
|
|
|
So it seems like most back up solutions are focused on backing up and, in the event of a crash, restoring the user's OS installed system. But I'm looking for something a bit different. I have an external HD with about 2TB of files that I've been collecting over the years. I'm tired of carrying the thing around and being afraid something's going to happen to it. I'm just looking for an online back up solution that will just store my old files online somewhere. I have three laptops I currently cycle through so it's not meant to be a back up of any of those or any other Windows installed systems I'm currently using, but rather just a back up of my 2TB of files on the external HD. What's the best software for this?
|
|
#
?
Dec 26, 2016 08:32
|
|
|
Any of the major players will do that, with some restrictions based on what OS you use. There may be some oddities based on if that drive is disconnected a lot or gets a new drive letter, but what you described is literally the point of consumer cloud backup. Crash plan, carbonite, back blaze, etc
|
|
#
?
Dec 26, 2016 10:47
|
|
|
-Blackadder- posted:. I'm tired of carrying the thing around Dropbox or Google Drive. If you want the drive backed up then to not have to carry the drive around I don't recommend Carbonite unless you follow some exact steps.
|
|
#
?
Dec 26, 2016 13:27
|
|
|
ConfusedUs posted:Crash plan, carbonite, back blaze, etc GreenBuckanneer posted:Dropbox or Google Drive. I've heard good things about Crashplan as well, but how accessible are the files once I've uploaded them to their server? Is it a hassle to get them off individually? Is it easy to make small additions to the files I have? I do like google drive (and already use it for a limited number of files) because of the ease of access and because it allows me to access my pdfs (among other files) without having re-downloading them (are there any other services that offer near the flexibility of Google Drive?) But there's no increment between their $10 for 1TB option and $100 for 10TB(more of a "business" option) and I only really need around 2TB. Not having smaller increment 2TB+ options seems to be the case with other big name cloud services like DropBox and OneDrive. I imagine most people don't have over 1TB of files to store so none of them feel the need to offer that option to anyone except businesses. I did notice that Amazon Cloud offers unlimited for $60 a year, that might be an option, anyone know if their service is any good? Or are there any other company's offering just above 2TB at a reasonable price with decent service? -Blackadder- fucked around with this message at 01:20 on Dec 27, 2016 |
|
#
?
Dec 27, 2016 01:17
|
|
|
Carbonite's personal and office plans (business) will delete missing files after 30/60 days of being missing or inaccessible (such as a hard drive slowly dying, or being disconnected) unless you "freeze" or pause the backup or don't have the software installed (virtually the same thing) The Carbonite server software you can set your own retention to as long as you want, but that's prohibitively expensive unless you're running an actual business. ($600 to $1000 per year) google drive/dropbox also have a 30 day (these files are missing for 30+ days, so we're deleting this poo poo), but if you're looking to drop something off onto someone else's server and not sync the data afterwards (maybe periodically connecting your HDD to sync every so often, rather than being connected all the time) then google drive or dropbox are some options. I don't know how their syncing options work too well if the drive is disconnected or whichever. I can't say too much about crashplan as I've not used it past a trial (backblaze as well), but you could potentially do what you're asking for with Carbonite. Just know it will obviously take time to upload over the net and then if you want to make sure the retention isn't going to kick in, to either stop the backup to prevent further changes until you next connect the drive (I say this because it sounds like you disconnect the drive to connect it to another of the laptops for a period of time, which would mark the files as "missing") or uninstall the software until your next sync time.
|
|
#
?
Dec 27, 2016 01:40
|
|
|
I don't like the suggestion for Dropbox or Onedrive as a backup solution. The single largest risk to your files is a crypto virus of some sort, where your files get encrypted. All of the solutions mentioned here will notice those files got changed/encrypted and happily upload the encrypted ones. The true backup solutions (crashplan, etc) will allow you to revert, en masse, to the pre-encrypted versions. You might have to call their support*, they can do it. The file sharing solutions (dropbox, google drive, etc) don't have that mass reversion option (unless that's changed in the last 6 months since I last did a deep look into it). *Carbonite can mass revert you if you call support, but Crashplan can do it natively in the client. I honestly prefer Crashplan. -Blackadder- posted:I've heard good things about Crashplan as well, but how accessible are the files once I've uploaded them to their server? Is it a hassle to get them off individually? Is it easy to make small additions to the files I have? All of these services will automatically back up any changes to the files you make. Download/restore is generally simple. I strongly suggest you just try a few to see which you like, but note that they may not play nice together if you have more than one installed simultaneously.
|
|
#
?
Dec 27, 2016 01:42
|
|
|
Crashplan's website will let you grab one file at a time, or bundle up many files as a zip file. The Crashplan client will let you restore any number of files to either the original location or somewhere else. I don't know about the home version, but Crashplan for business keeps near unlimited revisions of files including deleted ones.
|
|
#
?
Dec 27, 2016 19:53
|
|
|
Worth noting that Crashplan's Windows client is throttled but if you run the Linux version it is unthrottled so you can upload just as fast as your connection will allow. Even setting up a Linux VM just to run Crashplan would be a tremendous time saver on a reasonably sized data set.
|
|
#
?
Dec 30, 2016 18:11
|
|
|
I'm retiring my 6 year old RAID. Walmart has a 6TB Toshiba drive for $169 online, Microcenter price matched it. I'm on the fence between Google and Amazon for cloud backup. I know Microsoft's integrates with Windows really well, but they charge a bit more. Good point about the crypto viruses. FogHelmut fucked around with this message at 02:46 on Jan 1, 2017 |
|
#
?
Jan 1, 2017 02:42
|
|
|
FogHelmut posted:I'm retiring my 6 year old RAID. Walmart has a 6TB Toshiba drive for $169 online, Microcenter price matched it. I recently bought two 5tb external Seagate drives, and ripped one of the hdds out and put it in the tower to replace my failing drives. They were $115 a piece 
|
|
#
?
Jan 1, 2017 04:47
|
|
|
GreenBuckanneer posted:I recently bought two 5tb external Seagate drives, and ripped one of the hdds out and put it in the tower to replace my failing drives. They were $115 a piece Seagate broke my heart. I think I remember that Toshiba has Hitachi's old tooling, and the value is there at this price.
|
|
#
?
Jan 1, 2017 05:44
|
|
|
FogHelmut posted:Seagate broke my heart. I think I remember that Toshiba has Hitachi's old tooling, and the value is there at this price. Pretty sure Western digital bought up HGST.
|
|
#
?
Jan 1, 2017 18:23
|
|
|
redeyes posted:Pretty sure Western digital bought up HGST. Sorta unless something has changed https://hardforum.com/threads/are-todays-hitachi-3-5-wd-or-toshiba.1820956/ quote:Western Digital will be allowed to acquire Hitachi’s 2.5” and SSD businesses, but not the 3.5” business. Instead Western Digital will be selling that business to Toshiba – factories and all – along with granting licenses for the necessary patents, which would allow Toshiba to effectively continue in the 3.5” market from where Hitachi left off. This would firmly establish Seagate, Western Digital, and Toshiba as the 3 major players in the hard drive business across all product segments.
|
|
#
?
Jan 5, 2017 20:39
|
|
|
FogHelmut posted:Sorta unless something has changed https://hardforum.com/threads/are-todays-hitachi-3-5-wd-or-toshiba.1820956/ huh, thanks for that.
|
|
#
?
Jan 6, 2017 04:17
|
|
|
How viable would FastGlacier be as a cheap backup solution for my parents and the family photo album folders they keep?
|
|
#
?
Feb 5, 2017 11:45
|
|
|
No clue. I've never used it. Sorry bro.
|
|
#
?
Feb 5, 2017 17:23
|
|
|
The stars aligned and I had time AND space to run a back up then pick a file and try to retrieve that from the back up. It takes about an hour to complete a back up to tape of the whole file server, and then I got distracted and I'm not sure how long it took to read the file off the tape into scratch space on the old spare server but I'm guessing it can't have been more than an hour. I was able to mount the back up using Veeam and get whatever file I wanted and that part was rather quick. I'm feeling pretty confident that if something is on a tape, I can get it off again. Tape drive controller firmware update and Veeam update together seem to have fixed all my issues. My main problem now is that I'm only there on Wednesday and Friday afternoons so it's a bit awkward working out how to get an end of day tape backup working. There isn't anyone else there who I would trust to do it. So too bad I guess, twice a week will have to do.
|
|
#
?
Mar 7, 2017 19:10
|
|
|
Edit: I had a dumb question and solved it 30 seconds after hitting post, so for some content, what's the deal with Google Cloud Storage? I've got half a terabyte backed up there but have yet to be charged a cent despite doing a bajillion gets/puts/deletes over the last couple months (easily three-four terabytes of traffic worth). Do they ever start charging or what? The $300 credit expired long ago. Edit: by my math we did ~70 million Class A operations in one month alone, which should have been ~$350 in charges. Sheep fucked around with this message at 19:37 on Mar 8, 2017 |
|
#
?
Mar 8, 2017 04:57
|
|
|
Does anyone use vaultlogix? Just got thrown into supporting it and it seems to be having issues with a SQL AlwaysOn cluster. We have 2 nodes in the cluster and 7-8 groups, it doesn't seem to be very good at handling the backups for the cluster as it's failing backups when it's attempting to backup a database that is the secondary replica, which I kind of understand, but I have to configure it to look at all the databases because they could technically live on either node at the same time. I'm fairly certain I can't have it look at the SQL cluster itself and rather have to have it look at each node in the cluster. Anyone dealt with this have any suggestions?
|
|
#
?
Mar 24, 2017 20:30
|
|
|
Amazon deal of the day is an 8TB Seagate USB3 external drive for $184. Anyone have any experience with this line of external drives?
|
|
#
?
Mar 31, 2017 14:55
|
|
|
dakana posted:Amazon deal of the day is an 8TB Seagate USB3 external drive for $184. Anyone have any experience with this line of external drives? The 8TB and 6TB in that series are SMR drives. The 4TB may also be a SMR drive, not sure.
|
|
#
?
Mar 31, 2017 15:01
|
|
|
havenwaters posted:The 8TB and 6TB in that series are SMR drives. The 4TB may also be a SMR drive, not sure. Had to Google that -- I'm likely going to use this as an offsite backup for my photo archive, so I'm not too concerned about read/write performance as it'd hopefully be an "oh poo poo" backup anyway. Are there reliability or other problems I should be concerned about with SMR drives?
|
|
#
?
Mar 31, 2017 15:17
|
|
|
dakana posted:Had to Google that -- I'm likely going to use this as an offsite backup for my photo archive, so I'm not too concerned about read/write performance as it'd hopefully be an "oh poo poo" backup anyway. Are there reliability or other problems I should be concerned about with SMR drives? I guess it should be okay then? It's just slow and gets really slow if you try to write over a filled block again.
|
|
#
?
Mar 31, 2017 15:28
|
|
|
Sheep posted:Worth noting that Crashplan's Windows client is throttled but if you run the Linux version it is unthrottled so you can upload just as fast as your connection will allow. Even setting up a Linux VM just to run Crashplan would be a tremendous time saver on a reasonably sized data set. What would a transition from Windows to a Linux VM entail - stand up the VM, install the Crashplan client, mount the volumes in Linux and go? I am interested in doing this but don't want to re-upload everything that I've already pushed from the Windows client, and I'm afraid that would be required since Crashplan would see the VM as a new device...
|
|
#
?
Apr 18, 2017 17:27
|
|
|
At least for the business client they have a process to "adopt" a machine. I haven't done it in years so I can't offer any help.
|
|
#
?
Apr 18, 2017 18:48
|
|
|
havenwaters posted:The 8TB and 6TB in that series are SMR drives. The 4TB may also be a SMR drive, not sure. I have one. Its write performance goes as low as like 30MB/s.. not good. Reads are more normalish. It's a good value for backups but dont even attempt using it for workstation drive.
|
|
#
?
Apr 18, 2017 18:56
|
|
|
thiazi posted:What would a transition from Windows to a Linux VM entail - stand up the VM, install the Crashplan client, mount the volumes in Linux and go? I am interested in doing this but don't want to re-upload everything that I've already pushed from the Windows client, and I'm afraid that would be required since Crashplan would see the VM as a new device... I did this exact thing last year and I want to say that we didn't have any issues masquerading as an old machine that hadn't backed up in years but don't quote me on that. One thing you can do, if old data isn't going to be changing or updated (we store a lot of data that we will never, ever need again) is to just set the option to retain deleted files and then carry on with your day. We've got data stored in Crashplan that hasn't existed on our end in nearly a decade and everything is swell, I actually did a full restore just to verify things a while back and there were no issues.
|
|
#
?
Apr 19, 2017 04:17
|
|
|
Macrium Reflect question! We have a small business with 7 workstations (6x Windows 10 & 1x Windows XP, don't ask  ). The server is backed up nightly via Acronis. Data that isn't stored on the server gets backed up offsite via Crashplan. I would like to have semi-recent disk images of each workstation to minimize downtime just in case we do get hit by a Cryptolocker.At the moment I simply shut the machines down, boot a Clonezilla CD and clone images of each PC to a 3,5" HDD. Since some PCs are limited to USB 2.0 this process takes a few hours total so I only do it every 6 months, rotating two 3,5" drives, each then stored offsite at a different location. Would Macrium Reflect v7 Workstation be a better way of doing this or is it overkill? If I do set up a little NAS for centralised disk image backups via MSB, how do I keep a cryptolocker from encrypting all the Macrium backups? Would it be sufficient to firewall the NAS off and only allow access to it for a few hours per week (pretty sure pfsense has a feature for scheduled rules)?
|
|
#
?
May 15, 2017 16:48
|
|
|
|
| # ? Apr 18, 2024 03:24 |
|
|
I can't speak to Macrium itself, but in general you want a copy of your backups to be offsite and not actively connected to your network. There are plenty of backup suites out there now that will back up your devices to both local disk and to an offsite cloud location. Crypto viruses are getting smarter every day, so you really don't want your backups on the same network if you can help it at all. Nothing sucks more than finding out all your backups are encrypted, even those you thought were safe. Limiting access to your stored backups will reduce (but not eliminate) the chance of something getting in there and trashing them. You really want to have no access at all. For example, firewalls are only as good as the rules you create and don't necessarily protect against certain exploits.
|
|
#
?
May 15, 2017 17:20
|
|