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.
FranTech (BuyVM.net) loses all data - "FRIED NODE"
This discussion has been closed.
Comments
I don't think OP is even aware of the concept of a backup.
So, unlikely.
This thread reminds me I should probably backup my servers again... Keep putting it off but somehow I don't think that's a great idea
I don't think you quite get what was different here. We literally just used a riser instead on that one. What exactly is experimenting...? This isn't some new age idea, many hosts I know use risers for theirs since they don't want the IO bottleneck the onboard NVME slots give. Again, this node was in full production for nearly a year.
There's no API for it since then people would just script it and not pay for daily :P Making snapshots a manual thing is a compromise given we don't charge for them, no matter the size of the VM you're on.
Francisco
If you're on Linux, I'd recommend backupninja and Borgbackup. backupninja schedules the backups while Borgbackup actually does the backup. Both are available in the standard Debian repos (
apt install borgbackup backupninja
).Borgbackup backs up to a remote server, but most people here have at least one idling VPS they could use for backups, and there's usually very good storage VPS deals on Black Friday.
OP should've had his own backups, just for his own sake incase they delete something they shouldn't have. Still, I don't want to bust his chops or push blame his way, that's not fair.
The OP was offered 3 months of compensation and free nightly backups going forward. If OP feels that he deserves more for the heartache? He's welcome to propose it and we'll most likely accommodate it without issue.
Francisco
The bottom line is my Windows VPS with BuyVM that I've relied on for the past 10 months to use remote desktop all of a sudden stopped working.
Lesson learned. Backup your data. Move on with the offer Francisco has made you or find a new host. Simple as that.
Totally appreciate both perspectives, but coming back to reality. Nothing is 100% reliable and failure's do happen, on various scales - small or large (such as OVH... when their entire DC caught fire, even their own backup solution failed some of their clients in that scenario).
Ultimately, use the tools available to you - as Fran suggested the snapshots are free.
Sounds like this node was different, although it shouldn't have impacted the user in any way... most server setups aren't identical to be honest. Largely dependent on the individual engineer who configured the device. So, sounds like an unfortunate situation, take the compensation, schedule backups and move on. Should hopefully be covered in the future, with a backup in the unlikely event it happens again.
Happens in the hyperscalers all the time. They don't have live migrations. AWS doesn't click the snapshot button for you either. Their NVMe is instance storage rather than EBS and would lose your data the same.
More backups, less PMS.
More PMS, less backups.
(I want moer chaosssss.)
It takes me 20 minutes max to restore the most recent snapshot to VM with Windows OS.
Backup is a must if you would like to have a piece of mind.
Just download your disaster recovery plan. It will be fine.
Oh and you should disable PMS
Depends on the data that you need to back up. On most of my Windows machines, it's easier and cleaner to just backup the unique data and not the whole thing. Then you can reinstall the apps and just restore smallish config files and databases than 20-50GB snapshots from the same provider. When doing this, you can store daily/weekly/monthly copies since its smaller.
So far you just keep mentioning RDP, which is a basic feature enabled by any Windows reinstall and wouldn't need a backup at all.
Did the card blow up the drives or did the drives blow up their neighbors?
OP: This is why you keep backups.
Bad thinking. There is NO provider on this planet who really can offer 100% reliability - and particularly not at low or medium pricing levels.Nor is there any hardware manufacturer who can guarantee 100% working products for any period of time after and beginning with shipping.
Just burn it into your mind: Important data == make BACKUPS
I understand that you are frustrated, but again NO provider has magic 100% reliability powder. What @Francisco does that is, storing backups on another node/server is standard and good practice. As for "managed by BuyVM" you are (mis-)guided by emotions and blowing things out of proportion, actually BuyVM is among the most respected providers.
As you seem to be very concerned about your data, here's what you might want to do: get and use an additional second backup server/service somewhere else, so if your "main" backup ever fails you'll have another one. Btw. this is not specific wrt BuyVM but generally a good option.
@Francisco how about the shared hosting, does it have auto backup ?
OP, Upcloud offers 100% uptime. go with them if its really that important.
I would not recommend a 100% uptime host to someone that doesn't understand the concept of a backup. I would not trust a host with a 100% guarantee as there are too many devices between the node and the end user that are required to deliver the content.
@OP
Hosting and Toilet Paper are a lot alike. If you're taking a giant shit and are out of toilet paper you're in a tough spot so it's always recommended to have a backup.
i have 8 rolls of tp in my bathroom i can take shits all day no worries
cPanel does, DA doesn't since we haven't deployed Jet yet. I don't know if it's stable at the moment.
Shared (and probably resellers) will return under the Namecrane brand with a focus on high performance (think big EPYC/Threadripper CPU's), about the same user load out (but, again, these nodes will have 4x+ the CPU threads as the current ones), and higher resources (more shared cores for each user to help with Wordpress and such).
Price will be higher, I don't have exact figures yet, I'm still talking to the guys about that.
I don't know.
Neither drive is detecting at the moment in a different motherboard. Both were on the same card (bifurcation dealy). I don't know what caused it fully, just all of a sudden the node locked up and things weren't detecting.
I don't have any dumpster m.2 drives to gamble with to test the card further at the moment. I don't want to potentially waste a $300 NVMe on a $100 card. I'd rather just RMA board/card/drives. CPU/RAM is obviously OK.
For now I've spent a good bit of this weekend fiddling with things on a older workstation I have at the house with onboard NVMe slots. I got a few times where it would hang for a long time POSTing while trying to enumerate the PCI (which means it was at least sensing something in the slot), but never turned into anything successful.
Veeam is incremental so it works quick. I use it for some nodes and have set it up for some managed customers over the years. It works well and restores quickly and easily.
Francisco
Sounds like a shitty situation dude, hopefully you can figure it out. If it's got tits, tires or transistors it'll cause a man no end of grief.
NO BACKUP
NO PITY
I learned it the hard way, after one provider uploaded my files into the cloud
I got a Proxmox backup server after that, cannot complain
On a side note, maybe a nice VPS line would be a VPS with guaranteed backups (or > 99% guaranteed lol) - might not be a LET thing with the price though.
I've discussed in private a few times to include backups for 2GB+ plans, but I'd need to rack quite a bit of extra storage in each DC.
Our backups are pretty simple:
Problem is, there's no deduplication/incremental in that, meaning a user with a fully encrypted OS drive on a 32GB (640GB OS) VM, will generate 5 - 6TB of backups at any given time. It adds up quickly if we want to give 7 days of backups.
A compromise would probably be '1 weekly backup for free, 7 day backups cost the current rate'. Would still be chunky on disk space, but obviously much less.
Francisco
Like others have said, Francisco's bad on this one, but it's on you, too, for not taking backups. So he has apologized for his part and offered amends, it's up to you whether to take it or not. If you're too lazy to take your own backups, pay for it, it doesn't cost much. I've been paying $2.50 (or was it $2.00?)/mo. for automated backups on a Linode that I haven't logged into for years, simply because I'm too lazy to learn how to do my own backups.
>
Backups on the OP's plan would've been like $5 or so I think. Again, snapshots are free.
Francisco
Gotta love LET, expect a low cost VPS to have a zillion resources, 100% uptime, and no hardware failures of any kind ever. Sorry but even AWS/GCP/Azure won't give you the last point. On AWS, if your EC2 instance with local storage fails, you get nothing. I don't even think you'll get anything if your EBS goes up in smoke. If your data is worth anything to you, back it up somehow.
An uptime SLA means you get credited for downtime, not that it doesn't happen.
true that but they make sure it is very unlikely to happen:
its like buying insurance that you dont wanna cash out from.
of course their prices commensurate with that SLA as well.
thats still on the premise that OP doesnt want to do remote backups.