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
Looks ok here.
09:24:36 up 410 days, 17:41, 1 user, load average: 0.00, 0.00, 0.00Sucks. Shit happens.
Barring my experience with their support being slow, never had reliability/ performance issues with VMs over the 5 odd years that I have been with them.
Time to pull them socks up and work harder to prevent rolling further off the slippery slope @hosthatch
Yeah just disregard an email from HostHatch themselves saying a node in Amsterdam failed to no return
Yeah we're on different nodes
Surely OP has neighbors so it's just a matter of time until they drop by.
Yes, I also received this email.
"We ask that you restore your VM from backups at this time."
Maybe they can recover part of the data again? I mean HH claim they use RAID50/60 so I wonder what happened here.
As always lowenders: this is an LOW-END service with no SLA whatsoever, people's expectations for an low-end service is way too high these days.
Shit like this happens. Pay monkeys, get peanuts. Also:
I am in the same node too. I lost my VM twice. First in LA then this. All of them are storage. Don't know will stick with HostHatch storage after this incident.
Looks like HH is improving their disaster response. Some folks had to wait for weeks for a new VPS after a Chicago storage node blew up earlier this year. This time it looks like they recreated the VPSes quickly. Does not make the lost of data any better, but it is a step in the right direction concerning customer service.
unless providers specifically mentions , they take backup of data . You should have data backup always to avoid loss .Hardware failures do happen and it is very frustrating for providers and customers .
@miu , are you ok?
Ha, Amsterdam! Amsterdam is cursed, where is the Eindhoven/Utrecht location?
3months extension?
breaking my storage now!
are you doing it every 3 months ?
anyway they are good plex streamers, data dumps, go with others for anything persistent long term.
"testing backup recoverability strategy"
I agree. I think cheap storage is only useful for this. Lesson learned after previous fail.
TBF if they move you to a newer node it's a pretty decent compensation with the 3 months extension. It sucks to lose the data and you have to config the VPS but if you really care about the data you should follow the 3-2-1 rule.
Parity RAID folks!! Is x time more probable that RAID5/6(0) on large storage nodes will sooner lost another disks than able rebuild degraded array (than RAID10... minimally from reason that rebuild time is a much longer that R10 & performance in degraded state is too a much worse)
This is big reason why i always strongly prefer RAID10 when possible (not only better performance, but sure a much more safe). Yes, i know - so cheap and space saving solution (raid 5,6 based large capacity arrays) when many disks are used (only capacity of 1 or 2 disk provider loses for parity), BUT: serious probability that in troubles large capacity arrays will be not able rebuild himself sooner that will collapse total and lost all data. + often in time when 1 disk dies and fails, other are also on the end of their life and not able/realiable for survive difficult state, stresses and loads at parity array rebuild process
I was already on their latest Epyc storage nodes, which raises some questions on their claim in a previous thread that only old nodes are affected…
In the hours pre-death all the usual hallmarks of disk failure were present - freezing, random corruption, etc. There was no instantaneous data loss at first, was able to boot a few times during the issues
I assume that with one of CHicago Storage (EPYC) node is too not all good (where i have placed one vps), exactly the same hallmarks i seen last times there, and may be next on the order
Would you mind commenting on these recent involucrations @hosthatch?
I've always been a happy customer, but I think it's getting out of hand these days and a statement about what's the deal with all these incidents would go a long way.
Looks like I finally got a failed node
Panel says they sent an email over, but I didn't get anything. Luckily this VPS was a backup of a backup of a backup.
Yup, still reliable.
same problem 10TB of Data Lost
Even if a provider mentions backups, you should have off-provider measures. This protects from the unlikely but possible¹ event of a provider vanishing off the face of the earth without warning, the providers backup system not being well implemented and tested and found lacking when it is needed so not all data can be restored², the provider being very slow to rebuild infrastructure to restore to⁴, and so on.
[1] or maybe not that unlikely, with LE providers…
[2] do you even know what backup arrangement they have? Multiple copies or snapshots?³ How often they test it?
[3] if all they have is a very recent copy, that may be corrupt if a node has been slowly and silently corrupting data for a while before finally failing completely
[4] if that happens, and it has in some notable cases recently, having off-provider backups allows you to restore quickly to elsewhere instead of wasting time trying to convince them that your restore should be a priority
This is why multiple backup generations is important: if you only have a single copy, and that was refreshed soon before failure, there is a risk that these pre-complete-death failure patterns mean the only backup is corrupt.
Wait another location? Wowzer
Literally says in the title.
Chicago failed in the past.
Los Angeles failed in the past.
Now Amsterdam...
The end is nigh.