Europe's largest online storage provider, Humyo, has been offline all afternoon due to a multi-terabyte database rebuild. Humyo is an online, "store-your-data-in-the-cloud" service that was started up by Dan Conlon in January 2007. It has more than 300,000 customers using its facilities. Conlon funded it himself using money he …
a screw-up like that should kill the company
Not just because someone made a mistake and messed up the database, but because their architecture (god, I hate that word .... I should've said "design") didn't allow for a single point of failure.
If you're making a business of holding customers' data, it's *got* to be available ALL THE TIME. While outages due to external events - such as internet failures are a pain, anything internal to the business just reeks of amateurism. This shower should step out of the way and let the professionals do the job.
I'm going to make millions by building a device that guards against inevitable outages by downloading all the data from your cloud repository to a local disk. iCloudstore2.0 - $399 & $29.99 per month.
...doesn't sound nearly as impressive as it used to, given that my crappy home theater computer has 1.5tb online, not including network storage...
Show me a multi-PETABYTE database rebuild and we'll talk.
(the usual rant about not trusting other people with all your data)
Oh for heavens sake
I use Humyo. It is quite good, although the upload by adsl is tedious. It sounds like they had a bad experience and are doing everything they can to put it right. And, even you acknowledge, no data will be lost.
That's actually bloody clever of someone. Can you imagine any of the government systems managing to do that? in 60 to 120 minutes?
Look, it's not an iron lung. No-one died. It's not a missile shield or a near-earth-orbit asteroid database. It is where a lot of people hide slightly less than perfectly legal mp3 files, and their holiday snaps.
Why is it so absolutely, positively, definitely so essential that every bit of data in the world has to be available every existing microsecond? Libraries close at night and at weekends. The British museum is closed in the dark. Even the prime minister goes on holiday.
OK, so they fell down on some sales claim. Big deal. I never used it for anything where that mattered. In fact, I have always used it on the assumption they will go bust tomorrow and vanish in a puff of green smoke. And anyone who doesn't is a moron. Lets all get hysterical because the BBC made a claim that Johnathon Ross was funny, or because Golden Delicious apples are Greenish and Inedible. A few vorsprung durch technik motor cars go wrong. Are the rest of them useless as a consequence?
Its time to grow up. It's made by humans. It will be fallible. We do fallible oh so well
Paris, because she is fallible. And because you get less than it says on the box.
Just goes to show...
that cloud computing just feels like a major race to the bottom. If amazon's ec2 goes offline, a lot of businesses would topple and fail. And there is no guarantee that any cloud will stay online, of that if they have to go offline that they would do it nicely. Imagine what would happen if an overzealous Law Enforcement Agency seizes all of the servers in a cloud.
single point of failure.
I design expensive highly available systems (well, I used to before our entire division got laid off a few years ago, now I design really inexpensive highly available systems). The one thing I see in all this is that by using something in the cloud, most SMB are putting EVERYTHING into one basket: their (usually single sourced) internet connection. Then there's the inability to actually audit the practices of the company you're using to store your data - and the therefore inevitable downtime as someone, somewhere, fucks up royally.
Using the cloud as sort of a "hot backup" makes sense to me. Depending on "the cloud" to always be there, and using it to store time sensitive, mission critical data is setting yourself up for failure.
Anyone who uses a service provided (a) over the Internet; and (b) by a third party as their primary storage *deserves* to lose their data.
Humyo service details
Sent to me by Dan Conlon:-
Thanks for your call, we read your article.
Just to let you know - Part of our service is a desktop/laptop client. The software syncs users data between their local disk and the cloud storage.
This means that users using this software, still had access to all their data. Changes they have made to it will sync up to the cloud when they are next online.
The cloud will never be there 100% of the time, whether that be because of the user's Internet connection being down or our data centre being down. Equally, a user's local hard disk will not be there 100% of the time either - hard disks fail, computers get stolen.
So we think that our sync based solution brings the best of both worlds. Your local hard disk is there when the cloud is not and Humyo will be there for you when your local hard disk is not.
Robert is right
Robert E A Harvey is quite correct (and lots of examples too).
If your data is *important* to you and you really need to be able to have access to it at a time of your choosing, then store it on an external drive and carry it with you. An 8GB micro-SD card and a USB adapter can be yours for less than £20 on e-bay and these things are really small. (Yes I did buy one and it does work).
I also have 500GB of network hard drive connected to my home router and the instruction manual says I can set it up for FTP services (though I haven't bothered trying that). If you NEED access to your data, you HAVE to be in control of where it resides and how it is accessed. Equipment and services always fall over at some time, so plan for that.
Did little Bobby Tables log in ?
Every variety of storage fails at some point. The manufacturers of your failed HDD are unlikely to be making your data available again, and even if they can it's with charges and without guarantees.
Users were unable to access their data for a few hours in this case, which I understand is "less-than-optimal". However, the data was still there and now available again.
As the adage says, "Data that isn't backed up, is data you don't need", and anyone who uses online storage as their backup repository probably needs a better backup plan.
RE: Oh for heavens sake
Some of us are not human, we don't make mistakes. Please don't dirty all of us with your rotten views that inadequacy is OK. Its not!
There is a nice little saying that those who aspire for perfection use:
"A wise man learns by his mistakes, but a wiser man learns by others mistakes."
If you use this for everything you do in your life, you make surprisingly few mistakes.
If this bunch of monkeys did the same, understand, the Amazon Cloud and its chums makes mistakes rather regularly. Some of us know shit happens when it designed wrong, it doesn't happen if you invest properly in your plan and don't race it, unlike these and Amazon types.
Lets be fair, Cloud computing is simply wrong and a bad idea.
It is also billed as the primary source of your computing or data. The "stick it all here" type phrases are simply bad, misleading, stupido and asking for it.
Keep a local working copy, and "Sync it to this fluffy cloud" is a much better approach in this instance. But that wont sell will it!
Questions to Reg's Chris Mellor - or indeed any other informed source:
Does this vendor provide SLA agrements for their customers for the services(s) provided?
If so did this outage breach the required service level?
Secondly, does the vendor provide adequate support resources (staff, communications mechanisms) to handle the volume of interest and communication from their customers during a major outage?
What was the experience of users.. did they respond to emails and phone calls with timely, informative and reassuring messages during and after the outage?
I am constantly amazed at our (customers, potential customers) acceptance of very poor levels of service for "cloud services".
What happened to SLAs and decent support?
Why do service providers - from the "free" (ad-supported) through the low-cost and up to the expensive end - appear not to give a hoot about providing a decent level of customer service.
My experience has been that many of these orgs/companies just don't take customer support seriously.. and that many punters appear to accept the poor level of service as though it were "normal".
What do others think on this topic?
Ever tried emailing yahoo or google etc. and getting a human being to answer the question in a timely and useful way?
It is mostly a perception problem. just as people accept much higher risks by getting in their own car than they would on a train because *they* are in control, so it is with cloud vs local storage.
It seems their uptime is pretty good compared to to what you'd get with a local PC, plus you don't have to do all that security stuff yourself.
I'd have to agree though, if its just storage you're after, a USB flash drive sounds like quite a good alternative to the cloud.
- Boffins attempt to prove the UNIVERSE IS JUST A HOLOGRAM
- China building SUPERSONIC SUBMARINE that travels in a BUBBLE
- Review Raspberry Pi B+: PHWOAR, get a load of those pins
- Review Reg man looks through a Glass, darkly: Google's toy ploy or killer tech specs?
- MEN WANTED to satisfy town full of yearning BRAZILIAN HOTNESS