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.
we have lost our data with frantech
This discussion has been closed.
Comments
I stopped using frantech super long ago, they have too much issues for the clout they have here. So many better options for better pricing these days.
I will love some recommendations I making some testing with level7 or similar..
If it was not important, I guess you're just complaining about it because you're bored and got nothing better to do today.
If it was important, you would have a backup solution in place.
No, I've had a BuyVM VPS for 2 years now, and they've never lost any of my data. Apart from earlier this month, neither has there been any significant downtime.
vps are prone to data loss. why don't get a backup.
you can just buy another vps from same provider in another location and keep backups.
Sorry.
The slab node you were on locked up and required a reboot. When it came back XFS had a freak out and ate a bunch of volumes.
We don’t use XFS anymore but a few of the very first slab nodes still had it. These days we use either LVM or ZFS for slabs.
You can request credit and a rebuild. We will put you on a different node that isn’t using XFS.
Francisco
There is none. That's the funny thing. Everyone keeps saying they left, yet they always come back.
I currently run most stuff with German companies, best prices, uptime and overall quality, and pretty much have stopped looking for other providers so i might not have the best alternative list but some that worked well for me and i would recommend are:
avoro.eu: Even counting the dedicated resources thread from a while ago, i only have high praises for them and, the servers are performant, great uptime and affordable
I also have had very good experience with my.heartbeat-it.com previously my.server-factory.com, great pricing, uptime and everything, no complaints and opened 0 tickets with them after using it for 3 (or 2?) years
For servers outside of EU, I am not so sure of what to recommend as i have not purchased it in a while
I left few years ago (not sure about the exact date here either haha), and have never looked back, there was no reason to.
but sir how do i configure raid in my data
Hi,
get a provider who provides backup on different servers aside of the productive servers... or even better different datacenters....
Or, if you want to make it even harder for murphy, get backupspace from a 2nd, independent provider.
What i learned in the past years, there are no data that are not important.
If they are not important, you will simply not keep them or even pay someone to keep it.
So either just delete it or get a backup. Everything else is just lazy self-lying ;-)
this is true :-D
ALWAYS Backup all your important data other than origin server.
Use a managed service or hire someone to manage your servers, backups if you can't do it yourself.
but sir how do i configure raid in my data
We've had people buy 2 slabs and request both be on different nodes and then they do their own software RAID. Works fine, no issues.
Francisco
Highly recommend following the 3-2-1 backup rule.
Use RAILEP.
Mirror your data across an Redundant Array of Independent LowEnd Providers.
I use RAILEP-$7 for maximum protection.
I’ve seen that recommended a lot, but it just dosen’t seem practical. Maybe if you have a huge amount of data it is, otherwise, I never seen the point in 2 local copies and just 1 cloud copy for stuff that’s usually small like database dumps.
Might be something I’m missing though, as said, I’ve seen that recommended everywhere.
Personally I have set up our stuff to be backed up only remotely, Backblaze B2 & Cloudflare R2.
Of course, as said, probably would do it another way for large data.
I think it's not as complicated as some think it is. For ex, if you are backing up a laptop. One copy on the laptop, one copy on a nas or external ssd/hdd and one copy in the cloud.
Backblaze has a good blog post about it.
https://www.backblaze.com/blog/the-3-2-1-backup-strategy/
Yes, but for prod stuff you usually don’t run stuff in your home.
It’s not even sure that extra local storage is cheaper than cloud storage.
And it’s probably not true that extra network storage from your prod provider is cheaper than some cloud storage alternative.
And, to top it off, it increases vendor lock-in, what if your next prod provider don’t offer extra network mounted storage for a good price?
Idk, just seems complicated. Maybe one local backup on the same disk could be beneficial in most cases. I have a hard time seeing two be that though.
What are better options with better pricing these days?
Machine has been fine for 10 years and it's a system you didn't back up because it's not important...
But you want to leave the provider because you didnt have backups!?!
This seems like a you problem
Nah, that’s your interpretation of what he posted. People are so functionally illiterate nowaday.
Besides that, why were you so selective when you quoted him? It seems like you somehow managed to overlook the main part where he mentioned that he's just fed up with the constant issues he's having with a certain host.
Ya sure bud as he clearly says "it was a system that we did not have backup because it was not important, but the worst thing will be to recover the functionality of a system again that will be costly in hours of work without the data..."
Didn't backup will be costly I hours of work without the data. Having backup reduces the 2nd point.
On a side note if the system is not important, whys it matter if the data is lost... Oh wait it was important but they risked not backing it up and now they are paying the price for it.
So yes they are indeed wanting to leave the provider because they (the user) didn't create backup
Welcome to LET. Your providers will regularly face plant so it's advisable to keep copies of your data with at least two of them....and if that's not economic then there's always Hetzner and hope
My strategy for some services thus far
That’s it, I’m setting up backups for our Hetzner stuff as well
Its more or less a function of what people are willing to pay.
I have been toying with the idea of doing a trial runfor same KVM based services online, that i do for local clients.With optional upgrades for HA, Block Storage and backup intervals from weekly all the way down to hourly ( as well as offsite-replication levels like 1x/2x/3x). But i came to realize that it is just not really a thing for most users budgets on these forums.
Yes, Price can be an issue for some and depending on the provider, there may or may not be backup options. But, the point is to not trust one provider with your data. If it was for prod, having more backups is always good. That way you have multiple sources to recover data from. For personal use, having more than 1 copy is good enough.
Only had one issue with Frantech and that was back in the early days of slabs. I had a backup of the data elsewhere, so it was not a big deal. Fransisco identified the issue, then set me up on a new slab right away. He even offered some free service - but I never took up the offer.
To be fair, over the years, I have experienced issues at Vultr, Hetzner, Amazon EC2 and a couple other providers. And in reality, no system is perfect nor fail-proof, no matter the size of the company and what hardware and tech they use.
I use this same strategy with great success. I have RAILEP striped across storage VPS from multiple providers:
It's not like there ever was a case with 8+ providers were all actually run by the same people and would Deadpool all at the same time.
HXServers adds credibility
You know the worst part? It's literally the exact same issue. This LUX one was built around the same time as that so it was XFS, just like those.
It ruined an entire christmas for me. Instead of spending time with family I was trying to unfuck that.
I'm not trying to make excuses here. It was a fault, and users suffered for it. We've offered the OP a fix and some credit. I'm really sorry it happened, we've made changes to all newer hardware many years ago address it, but rebuilding the older nodes isn't easy given you're pivoting around 130 - 150T at a time.
@ediaz Please let me know if there's anything I can do to help make recovery easier.
Francisco