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
Nah. @Francisco is going deliberately down-market.
Has anyone had an update this morning???
no, nothing new to report yet. I honestly can feel your pain, losing important stuff without proper backups will keep your heartbeat high for quite some time
I know that words and suggestions won't help much in your situation right now - but try to take this as an opportunity to rethink your backup strategy and such.
Used to live without much backups once... now I do something like three different backups from every important production stuff, each to different location/providers and so on.
would like to see their bandwidth graph after it comes back online as everyone tries to get their data.
thanks for understand, and yes, this is the worst situation, i suppose its my fault, right now i only want to access my data before monday(14 hours countdown here), but im concerned because @AshleyUk dont connect here since 24 hours, zxhost site is down(only billing works), no ticket answers, no skype answer, and 4 days already...
May God help me, if i get over this without all my clients gone i will make backups in all over the world.
thanks again for understand.
May God help me, if i get over this without all my clients gone i will make backups in all over the world.
Yeah now the horse has bolted
is not like that, i have two vms with ZXhost, server and backup(yes i know is not good practice), but the vms where in their 2 different DCs, last week they moved one of them to the same DC of the other, i was waiting for black friday to see if i find some interesting alternative backup place.
last ticket(little downtime) they calm me down saying the enviroment was stable for months, and suddenly this happens.
Give @AshleyUK some time I'm sure he is just as unhappy about the turn of events as you are.
Anyway while I do not think it helps much to ask them for an apporpriate compensation (not just what you agreed in their ToS for the SLA), you can always try that. How it works out to sue a company which is not in your own country..usually not worth the costs. Anyway in Germany e.g. a provider claimed 99% SLA via ToS but a customer provided proof that he had 2 downtimes (not even that long I think) and that he would have made considerable income in that time and he was given right. The court nullified that 99% SLA part of the ToS and stated that the webhosting was the main part of the contract to be fullfilled and as it was not the provider still had to compensate.. far more than what you'd get from SLA.
"Doch selbst, wenn die AGB eine Einschränkung der Verfügbarkeit vorsehen, ist diese längst nicht in allen Fällen wirksam. Das LG Karlsruhe hatte im Jahr 2007 einen Streitfall zu entscheiden, bei dem der Server eines Kunden zweimal vorübergehend nicht mehr aufrufbar war. Dem Kunden entstanden deshalb enorme Umsatzeinbußen, die er von dem Webhost ersetzt verlangte. Der Webhost berief sich jedoch auf eine Klausel seiner AGB, wonach dieser eine Erreichbarkeit seines Servers „von 99% im Jahresmittel“ gewährleistet.
Diese Klausel war nach Auffassung des LG Karlsruhe unwirksam. Grund dafür war, dass die Erreichbarkeit des Servers nach dem Webhosting-Vertrag die Hauptleistungspflicht darstellte. Der Ausschluss der Haftung auch bei schuldhafter Verletzung solcher Hauptleistungspflichten ist jedoch nicht möglich. Das LG Karlsruhe sprach dem Kunden deshalb einen satten Schadensersatzanspruch von 97.201,71 € zu."
Anyway let's not be a dick and wait for Ashley to fix things
one needs to add that the Landgericht is not the highest level in jurisdiction here and also know that the same case may have been judged completly different by a Landgericht in another federal state of germany... ;-)
True that. I'm by no means a lawyer and the various courts argue and make decisions differently of course. Also the LG is - as @Falzo pointed out - not the highest level of jurisdiction here. Just wanted to point this out I dislike how everything written in ToS often is considered the "holy word" where as it is not. Not necessarily.
People complain when a host goes dead without a notice.
People complain when a host gives out 60 days of notice.
Conclusion: Just go deadpooled without giving a notice at all. The end result is the same.
You can find a glimpse of what is going on behind the scene on the ceph users mailing list
Search for Ashley .... Thread called OSD Random Failure
.... so they are working the problem and probably not getting too much sleep.
Well, I know I'm keeping the hell away from Ceph.
Yes - I had CEPH planned for a small university cluster - now I'm looking for simpler alternatives.
They do seem to exist - so far come across "Sheepdog" and "LizardFS" ...
... so there might be alternative uses for Ashley's hardwre.
Same here. Never liked ceph tbh.
How many of you are getting the dwindling feeling that ZXHost might not be able to recover the data? and might not even come back online EVER for their storage servers.
I am awaiting that dreadful feeling where they might come back and say, sorry but EOL is here already and here is your pro-rated refund (if they decide to still provide the refund).
I would find that pretty shocking/dumb/etc if Ceph really requires ever single OSD to be online, even when there's replication enabled. Erasure encoding is only now supported for RBD/block devices in the latest bleeding edge stuff. Even then, the performance is bound to be so garbage that no one should be using it.
My recommendation to Ashley would be to check that the drives that are problematic are healthy. It's possible it's hitting a bad sector/etc and causing it to crap out. If the drive really does have a few bummed sectors, a dd_rescue could get the data (minus those few bummed sectors) mirrored over to a fresh drive.
Francisco
it is at least a case that's possible and one should be aware of. remaining question would be how long would one expect Ashley to try and fix things or how fast one would expect Ashley to cut it in favour of what's left beside storage VMs.
I can see high latency on my other services still running there, like:
maybe congestion on the switch or something like that, so probably a lot going on there internally. it's been like that for a while (>24h), sadly I only noticed because I have that on services I use for latency critical stuff
@Falzo if a 1.3% packet loss with 200ms is breaking your shit- you need to redesign your shit.
I didn't say it broke any shit. I just said I noticed because latency matters in my usecase of those specific VMs in frankfurt... (you know there should be fallbacks for mission critical stuff and such ;-))
Ok, fine. I'll rephrase. if a 1.3% packet loss with 200ms is causing enough loss that your setup isn't working well enough- you need to redesign your shit.
yeah, there is not much redesign in having a frankfurt VM to access other services in frankfurt with very low latency, don't you agree?
again, I am not complaining at all or saying I have a problem here, I have other VMs in Fra working just fine :-)
Only if they're on the same network. Otherwise, you're just biding time.
*sigh*
hope this means he is doing something to restore files. and can be have the things up before tomorrow 8 am.
the silence here, in emails, tickets, private messages, skype, network status... is killing me. if he cannot give back the vms in 4 days, at leats he can try to restore critical files from clients like me, i offer him to give the file names, no more than 15 that i need from any of my vms with him, but nothing, i got 0 responses. absolute silence, not even readed.
Why should Ashley make an exception?
Did you pay extra for that "we'll suck your careless maggot-infested dick within minutes 24/7" option?
This here is a mistake, it's a mistake for you to put your own ETA's because they won't be met.
It's weekend now and Ashley is depending on the developers of Ceph aswell to review the issue - I wouldn't think they work during weekends.
So no you shouldn't wait nor plan for the issue to be solved by 8am tomorrow morning.
You should have a contigency plan yourself, if you don't have one then that is your fault and you cannot hold anyone accountable other than you and if you haven't done so already you should inform your clients about the possible downtime that will be involved.
As for the silence, again its weekend - if there is nothing to update you then what should he update you? And no Ahsley clearly mentioned that he cannot put any VM's back online as that destabilizes and puts the rest of the VM's at risk - so why should he risk all his clients for you?
Just some food for you to think and plan as yes tomorrow will be a long day for you and you need to understand and accept that.
I got an email this morning, saying the same stuff posted earlier in the thread.
I was interested in 10EUR/year service at 1Fichier, but don't have BTC. Unfortunately, that rate only applies if you pay by BTC . Also, can you download your own files through the FTP or only upload them and you have to download through their website? Is there any way to upload via commandline/VPS?