All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Why I'm sticking with VPS6.net
I have had service with VPS6.net for about a month now and they've been rock solid.
However, a couple hours ago, I had the unfortunate problem of having my VPS completely wiped. This was accidental of course, and I was not mad because they were honest and professional about the problem. I just sighed... half a month of work gone...
You can probably see where this is going... I cannot stress how important this is: BACKUP BACKUP BACKUP. I was just about to implement a backup system when my server failed. (Really, I was in the process of creating a duplicity backup system.)
But the reason I made this post is to share that, although VPS6.net did something pretty crappy, they made up for it and were helpful and clear in getting it resolved. I've learned my lesson; now I'm going to always keep mirror copies of the same server...
(Although, I guess the $20/month for a Linode server is a better guarantee that it won't be accidentally rewritten by a tech support rep.)
Comments
Ouch. I'm sorry to hear that. But, the bright side is, it won't take you as long to setup what you already did
Of course, there's nothing stopping a rogue Linode admin from doing it intentionally, or their internal control panel from getting hacked...
@NickM: Yes, but Linode has a native backup add-on, which I have been using and which makes it nearly impossible for you to lose data. I guess it's true that nothing stops this from happening on Linode, but I think the 4 snapshot system will prevent you from losing any important data.
That's a good way to look at it. Although I did have a couple recent web designs that I lost.
If I was a rogue admin and I decided to destroy customer VPSes, I'd probably be thorough and destroy their backups, too.
Backing up to a completely separate provider at least prevents that, though of course if your VPS is compromised, they may very well discover the keys/passwords needed to connect to your other provider and...
So what happened, you said the vps failed? Did the node crash and they lost your data? did the support crew shutdown your vps and wipe it?
Inquiring minds (and fellow VPS6 customers) want to know...
What happened: I signed up for another Xen server, and went on live chat to ask when it would be deployed. The customer service rep said it would be about 20 mins. I waited half an hour and logged onto my SolusVM, only to see that my happily-running server now had CentOS installed (instead of the Debian it was running on). This is weird, I thought. After refreshing the page and attempting to logon using ssh, it occurred to me that the server was actually wiped clean.
I asked frantically what was going on. The customer service rep had an moment to realize he had just reinstalled my painstakingly-configured server... he said that it was his error while he was deploying the server, he thought the server I was already running was actually the new one, and accidentally reinstalled the machine.
But no, there was no node shutdown or VPS failure. It was accidental human error.
I can actually respect a company that is willing to tell their customer "We fucked up" rather than what most hosts would have done in the same situation - blame it on a bug in SolusVM or something of that nature.
You should never ever ever ever use a host's backup solution as your only backup. There are faaaar too many things that could go wrong
@bill32 Thanks very much for your patience and understanding. If there's anything we can do to help you get back on track as quickly as possible, please don't hesitate to let me know!
Was it Chris who you were working with on live chat? I'm definitely going to see about this, because he shouldn't be doing things like that...
Hopefully Chris offered a refund, at least. If not, I think that would be in order.
I should add that we are developing a system like this as well, to be available as an add-on for Xen-based VMs. We will be deploying a SAN to host backups sometime next week, with hopes to fully launch the service near the end of April.
If you lost valuable data, it's your own damn fault. Sure, it's a pain in the ass to reconfigure a server (though important and/or difficult configurations should also be backed up), it's not the end of the world if you have backups.
Sounds like organization in your company is lacking..
Yes, Chris did offer a refund. VPS6 has been pretty good except for this incident, so I decided to try again from scratch and keep using their services. Instead of a refund, I saw the mention of the VPS6's new backup system and decided it was fair to get this service for free for the lifetime of my servers at VPS6.
The fact is that this problem is not a recurring hardware bug or bad server configuration. It was a mistake. And I recognize that most customers will never experience a mistake as big as this. Although this was really crappy, I'm willing to overlook this for VPS6's quality infrastructure and rather amazing value.
I have two encrypted daily backups going on my Linode server (which I plan to eventually move away from): one to Hostigation and one to Amazon S3. Unfortunately my backup script didn't want to work, so I put it off (for too long). Now I'm always going to mandate this backup system in place for any server I have longer than a day.
If that were to happen to me, I would never go back to that company. 'Nuff said.
Shows a lack of professionalism to say the least.
@Infinity: Even professionals makes mistakes sometimes, but I definitely get where you're coming from- this should not have happened. I'd say VPS6 is lucky they have such an easy-going and understanding customer. I have two different sets of backups for my stuff, so the data is in three different locations, but, I wouldn't personally tolerate a mistake like this, especially considering things like databases that you can't constantly have a perfect backup of.
@bill32: Best of luck getting everything back up to the way you had it. And back up everything. :P
The real thing I don't understand is why you were developing websites on the server and not have a local copy. I do dev, design, and server work for a living. I've got backups of backups. It is the absolute one thing that you need to do is backup data.
For one, you can't really go wrong with Linode.com. They have a fair and balanced TOS as well, and their support staff is really good. The other thing that I can tell you is that it is absolutely unacceptable for a hosting company to delete a VPS. Unless some sort of catastrophic failure occurs where the drives on a node go bad and data becomes so badly corrupted that it's beyond recovery, I don't see any good reason why a web host should erase "by accident" a customer's VPS. Backup or no backup, slip ups like this are bad, and can become a PR nightmare.
I agree completely. Please don't think I am anything but extremely concerned about this -- this kind of issue has never affected one of our clients before, which makes this doubly concerning. It is my every intention to ensure that nothing like this can happen again.
I am only glad that @bill32 has had the patience to see things through with us in a cooperative manner, despite the terrible mistake on our part.
@vps6net I haven't actually looked at your website, but now I did and saw that you are using SolusVM for VPS management. I can actually see how that can happen, and while a customer may put the blame on you, I blame SolusVM more than anything. It is extremely easy to delete a VPS in SolusVM if you have Administrator privileges. One stop gap measure would be to create multi level accounts for administration where certain administrator accounts have certain privileges, and make it extremely hard for an Admin to remove a VPS. The same thing can happen from within WHMCS, if you have admin privileges, one click on "Terminate" and it's gone. Now, of course all that should be transparent to a customer, but the sad reality is that only the biggest hosting companies can afford to develop their software in house. Most of the software that companies develop in house for VPS management and billing is written in Ruby on Rails, and they are using crude API calls via CURL. Anyway, I hope that the issue will be resolved to the customer's satisfaction, you probably know how to best handle this
I wish you the best, and I hope that commercial software makers like Solus Labs and WHMCS will implement more safeguards in their applications.
@Dionysus I guess because people love to complain...
To let us know he was sticking with them and that they're that awesome.
Or, get the hardcore version: You broke it !
Yeah, being honest is really important for me. I dont mind errors, I had my fair share over the years, including things as deleting in the wrong panel of MC, for instance...
If they do break it, then I have back-ups of important data, if I dont, well, guess who's fault is that... However, some apologies will help the situation, even when nothing can be done to recover something.
M
Another problem you will want to watch for is provisioning multiple orders simultaneously. Solus has no queue system for provisioning, and we saw over and over multiple VMs be assigned the same ctid, ip, etc. If you don't catch it right away, then terminate one of the two duplicates later on, the other poor guy is affected as well.
@vps6net - If you stick with Solus, I strongly recommend doing frequent audits for duplicate data on your vservers table, as well as across your nodes.