All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Hostsolutions' vel7 (storage deal) down for > 1mo, no tk response
I know some people here are tired of seeing hostsolution threads, but there aren't any still open thread about this ongoing outage, and the provider is not answering the ticket, so here I am. I have another vps with them, and I'm completely happy with it, so I hate to write something like this... but I can't find another way to get some reponse. I'm not trying to pick a fight, I just need to know what the situation is to make my decisions.
It's been more than a month since this happened, and vel7 (and probably vel8?) is still down.
Then I asked using a ticket to wipe my data out and move me to another node about two weeks ago(June 25th). It got unanswered and is closed now.
The notice here is the last one regarding this problem. I haven't heard from them again regarding this node's being down.
At this point, I have even come to think @cociu has found out this storage deal is unprofitable and that he is deliberately trying to make people take refunds and shrink the nodes.
to @cociu:
1) Why would a simple failure take more than a month to recover? Did the drives fail completely and you're waiting for RMA's to replace them?
Why are you not providing any details to the customer who are waiting because of your service failure?
And more importantly, when do you think the node is going to be online?
2) You said you're moving people to other nodes for those who ask for it,
cociu said: 4) we have give the posibility to all this customers to ask for a new vps and we will provide in order what we have recived this ...
so why did you just ignore my ticket and not even reply with something like, "the other nodes are also full, sorry" ?
Comments
Another thread. Another drama. Make a new domain LowEndTalkAboutHostingsolution
Funny. I always get an answer when I open a ticket there; usually even quite quickly.
The one provider here whom I found to have lousy support (waiting long for an answer which then usually is canned and/or clueless) was Aruba. But hey, what I get for my 1€/mo keeps me happy anyway.
@cociu
How about finally getting those fucking Vel nodes working?
You really don't do a favour to yourself allowing ever new complaints creeping up. From what I see you went to fast with those storage VPSs plus you had the bad luck of badly timed power problems. No problem, we all make errors from time to time and bad luck gets everyone sometimes.
And compliments for always being honest about it and for obviously trying hard to get things working.
But this bleeding should be stopped, no matter what it takes. Throw refunds at unhappy storage customers, put them on another, on a working node, whatever - but stop the bleeding!
They will give you a refund. I ordered in April and never got a working VPS, so I gave up on the wait. @cociu is a great guy, but I'll never buy from him again due to the cluster f@ck vel7
@cociu any coupons for this promo ad?
That was also the case for me for other tickets.
I'd be careful with that. Yes, quite some information seems to point to complaint threads in the end bringing business. However: what we do not see is all the business lost due to potential customers reading such threads and staying away. That might be a small number but it might also be a quite big number.
I know, that's why I waited for more than a month for the repair, and about two weeks for the ticket reply before complaining publicly.
To avoid misunderstandings: I did not in any way attack you! My last comment was a response to maldovia.
oops, sorry
i am tired yet , please let me know what you prefere , full refound or a vps ? Pm in private your account .
I will be short in answer , nobody was ignore you , we have take vm to vm and in the first 2 weeks we have wait for fsck. After first 2 weeks we have start to provide new vms to each customer andf add +3 month for compensation . Unfortunatly i dont know what happened with your account. Right now we have to deploy only 10 account more from a total of 67. I know you are not in this list because i have talk with all this 10 users. So is possible to be a error or eaven we have deploy your new vps and you have not recive the email or something. Nobody want your money no worry. like i tell you before , pm and i will resolve it today.
Also the vel7 is still fsck (yes is unbeleve , have +3 weeks now)
Also i have a personal question , like you have opened this thread blame my company , way you not have try to pm tell something like this " i have opened a fuking tiket idiot like xxxx days and nobody is answering can you help me ?" and i am sure this poroblem was resolved before... any way is your right.
Thanks for reply, pm'd
just to update , to proof this vps was delivered many days ago here i have the printscreen with the client account http://imgur.com/a/t6hvc . Now we have close the tiket beleve the client was recived his email with the new details. Unfortunatly seems not recived.
Well if you actually responded to the tickets you wouldn't have any of this nonsense. Instead of ignoring tickets and saying you're busy, why not go to all the users and let them know you are currently busy. Pretty sure canned responses aren't hard to setup in WHMCS.
Everything resolved well, thanks for quick reply @cociu.
you really should think about the technical design of your storage nodes. you once claimed to have 70 of those...
if that is really the case everyone should take a minute to think about the risk and likeliness of another harddisk failing in any of those nodes and how this might affect all other customers on the nodes again because of week-long fsck, relocating VMs causing shrinking and so on.
I think people here like you @cociu and that's why they care and try to help you with fair and good advise. while sometimes sounding a bit rough those most likely aren't meant to offend or complain.
so I absolutely agree with the advise of giving communication with customers much more priority... if you'd follow this I am pretty sure as an outcome you will see much more time saving by investing soem work in replying friendly to tickets, mails and whatnot instead of trying to save that time by simply ignoring them.
Had a new VM in another node without asking anything a couple of weeks ago.
Works great since then. You'll get that fixed @aeba
There's a reason people don't do large scale storage on OpenVZ because of crap like this.
At a certain point you're simply not going to have enough memory for FSCK to run, even if you got 128GB in there. While EXT4 can support more than that, no one in their right mind is going to want to store that much in a single spot.
I've never seen anyone claim an ID higher than...10? 20 maybe? I've never seen anyone claim to be on a node in an ID any higher than that. If he hosts his servers in the same AS as his website, (possible given this ASN has Liberty, etc) thenit only has 1.5k IP's **total*.
I dunno. It sucks, poor @cociu probably looks like Homer at this point.
Francisco
Honestly, if I was in a position of waiting 3 weeks (much less in reality) for an fsck, I would simply refund everyone, decide it was not a viable business model, wipe and sell the physical hardware to recoup as much loss as possible.
Completely ridiculous.
When it was first announced that the node would be going for a FSCK I knew it wasn't going to come back. I've tried to do large scale fsck's like that on BTRFS, EXT4, etc, and none of them completed.
I let BTRFS run for like 3 days before I just nuke/paved it to ZFS and left it alone.
Francisco
indeed, frankly, I don't believe that anyone would let a fsck run for 3 weeks, it is ridiculous, I think there is some other issue here.
Well, hopefully fsck doesn't happen on the working ones.
I agree, @cociu self claimed it to be at least 70...
https://www.lowendtalk.com/discussion/comment/2232265/#Comment_2232265
maybe he confused nodes with harddisks though...
It's always possible that it's as I said: he just has a few very rotten eggs making his life miserable.
Still, storage plans on OpenVZ are a big. fat. no no. Use KVM because at least then users can self FSCK or in some cases they don't need it if they weren't in the middle of writing anything.
He wouldn't have the shrinkage issue either but I think that might be self inflicted.
Francisco
While I agree that doing it on ovz is a nightmare (well, what on ovz isn't?) I do neither think that @cociu should just end it nor that he will do that.
As I see cociu (and I clearly say that I like that) he's not the kind of guy to easily give up. If he's decided to do storage VPS he'll go for it no matter what he has to endure and go through.
And it is feasible (although, clearly going the ovz route wasn't the smartest approach). Given one has the resources and at the right price (which I have reason to assume being the case with cociu) one can do it - and - earn money.
As for the fsck - fuck, what's the alternative? Let the user data burn down and say "sorry, tough luck"?
That's the thing. With such a large filesystem, and it was a dirty shut down, there's a super high chance that if the FSCK ever starts running it's going to have a wall of multi-assigned blocks or things like that corrupting everyones data.
Ploop doesn't help with this much since it still sits on top of a main filesystem. If anything bugs out on that (bad fsck, etc) there's a good chance things are going to end very badly too.
Francisco
Right, but that's also the point in using OpenVZ: you can oversell like crazy. Long fsck is the price to pay...
Oh well, @Francisco, I guess we both know someone who has learned (among others) the lesson of good hardware raid w/bbu...
I'm wondering anyway, btw, that quite few (it seems to me) providers have a serious backup backend. Considering how cheap 10 Gb switches and cards (better halfway current servers have them on board anyway) have become and the seriously fat - and affordable - hard drives that are available I (if I were a provider) would certainly have one or more 100+ TB backup boxes connected via dual 10 Gb links. And it would easily pay for itself due to plenty customers who'd gladly pay 1$ or 2$ more for good and fast backup.
Yep, 3 - 4 weeks to do a disk check, common.
They are backup servers, after that long your backup is no longer relevant and no doubt you have a more recent one elsewhere if you are using it for Linux ISO's you can redownload them.
If I can't get it back after 3 days a week at the absolute most depending on circumstances then I am booting into sysrescue, turning off the fsck, backing up the container config files and reprovisioning from scratch, you will be back up and running within 4 - 7 days instead of 3 to 4 weeks a few times a year.
You would think so, I have been offering it for free for 7 years, less than 50 customers have taken me up on it.
We're talking about backing up a bargain basement terabyte VPS. That won't happen for free. I have to wonder what the $1 or $2 more is supposed to mean: per TB? Not realistic either. I generally only use this type of VPS as backup for other systems, rather than as live servers. That's just me though.