Maybe it was a small cluster of systems that failed which was taking more load then they thought.
This cloud computibg is BS, it all depends on hardware when said and done.
Google yesterday apologised to Gmail users whose email was out of action for more than a day. The world’s largest ad broker confirmed in a Google Apps discussion forum late Thursday it was restoring the service for a “small number” of peeps affected by the outage. Unsurprisingly, it didn’t reveal how many people were unable …
Cheap application of some kind doesn't have five-nines uptime.
If you don't want the trouble of installing, configuring, patching, feeding, watering, developing and otherwise maintaining your own global e-mail infrastructure, you either pay for it or get it for free. Either way, you get what you pay for.
Last I looked, Google Apps was quite explicit about exactly what guarantees you get for your money (free or premier). The fact is, the occasional glitches pale into insignificance as compared to the cost of running the infrastructure yourself for many people. You don't get something for nothing, though.
If you don't own the cloud, and have no (contractually?) binding SLA with the owner of said cloud, do not, I repeat, DO NOT, put mission critical systems into that cloud.
I imagine it looks cheap but you need to assess the risks you are taking on. I probably wouldn't factor in a risk that Google might go out of business, but both Google and Amazon have had enough cloud outages that, had I any of my customer facing or process critical systems deployed there, I would be looking to move them, quickly.
You get what you pay for. If you want five-nines uptime you will pay for it. Whether you run it yourself or trust to Google or Amazon to do it for you.
Mine's the one with the purchase order for a pile of blade servers in the pocket.
A days down time is nothing - you should try working in the environment I have to suffer. Naturally it's a MS Outlook/Exchange environment which is maintained for our organisation (a local authority) by a well known outsourcer and it's down/broken more often than it works! My private gmail account knocks spots off my corporate email system by a mile
Actually, Google's pretty damn reliable and unless you have an SLA contract that guarantees the mythical 'five nines' quit whingeing. Good luck building your own alternative, too (not because it's especially hard but rather because anyone stupid enough to use a Beta as mission critical and then complain about a *day* of downtime, is not going to be up to it).
I have a premium google email service and only had 1 small inconvenience. I suspected a problem but didn't have any specifics until the El Reg put it up. Thanks for all the good work.
Memo to the hard charging corporates worried about 5- 9's...
Whats the use when the jerks on the receiving end say they never got the email anyway?
Then you look in their inbox ( I have remote access to my users), and its sitting very plainly, unopened- in Trash!
Email down-= Time for a coffee break or dash into the pub for a while in my book!
The UK monarchy is in line for a share of blame for the Gmail downtime, I see (above).
For shame, you guys!
It's more likely to be US lobbying (backstabbing?) overload.
Sure, Gmail is generally a reliable application, anyone like me who has got an original UK 'gmail' tag (as opposed to the newbie 'googlemail') must have found it good enough to stick with, but that's all the more reason why we feel the loss when it fails. But, equally, I agree, I wouldn't run any significant business from a gmail address.
BTW, how long has gmail been on 'beta' now?
BTW, still locked out.
Does anybody have any good news?
I doubt anyone can make a 'cloud' that can be trusted for mission-critical applications. Such as using it for all your word processing/spreadsheets/etc. Reminds me of the companies that used only Skype for their telephone connections, and were crippled when it went down for several days.
@ Bill Gould: Think using gmail is bad? I know one company whose email addresses are @aol.com .
I just want to put my two cents in that the whining by sys admins etc. is just bogus. I use gmail for my domain (free version) and I didn't have an any serious outage this last week, and maybe I would be whining if I did, but its just silly. It is the sys admins afraid for their jobs -- don't worry there will always be stupid or lazy users for to serve.
Maybe the gmail team was caught asleep at the wheel, but I know what it is like inside -- I used to work there -- and I can guarantee with or without this whining they will very quickly fix this problem and address any deficiencies for premier customers. They will do this so much faster then the evil empire to the north has ever solved a problem, any problem, as to make this discussion a joke! You have to remember that this is what that company is using inside its walls and they did for a year before it became a product. If you have seen G from the inside you have no doubt that it is already fixed. And even though it is already fixed they will make it better and more robust in the next two months because of this outage.
And far as reliability goes, hell any company's Internet access goes down more often than gmail will ever go down. If you have an issue with gmail the issue should be continued internal access even when your Internet access goes bye-bye. But again that is the thing, the reliability of the infrastructure used by Google is several nines greater than reliability of any single company's infrastructure, so this discussion is, as I said, bogus and self serving.
OH GOD...as someone who's been with AOL as an ISP since c.1998, and seen what used to be a pretty good service go downhill rapidly with the outsourcing of customer support and the takeover by CPW of the UK group...I'm looking for a route out....fast.
My own SMTP server on my own box going out via DynDNS MailHop is more reliable than their SMTP!
Speeds dropping like a stone....
Average of 20 outages a month on the DSL connection...
Anyone suggest a decent ISP with no thottling, uncapped (at least 60GB/mo) and no port 25 blocking? .. oh and maybe an uptime/speed SLA?
""This is not a temporary problem if it lasts this long. It is frustrating to not be able to expedite these issues"
Ohhh, did the big mean Google cloud not give you your mail for a whole day? Ohhhh dear. As others have pointed out, you get what you pay for. If you have a big enough organisation that you need mail clustering with full site failover capability, then you'd better implement it. Or purchase real services from real businesses like Fusemail that offer failover and 5 nines support.
(PS, I use Gmail as my primary *personal* mail addy, but I'm not going to run a business using it)
It surprises me that anyone would expect email to have a 100% reliability rate from any provider. Google is certainly no worse than most and considerably better than many others. I have repeatedly had to tell customers and users that email and text messages while usually quick cannot be guaranteed to be instant or even arrive at all. If it is mission critical then call or do it face to face. Emails and text are a convenience and should be treated as such. (expect to be flamed, duck and cover)
Biting the hand that feeds IT © 1998–2021