With such vital services and heavily used down...
...you'd think there'd be a lot more activity in the comments here.
(Sadly, I'm lost without my Google Talk, but only because it's my daytime link to my wife.)
Three Google services were gripped by disruption today – just four days after the company dramatically reduced prices for one of its online storage lockers. The web giant's online instant-messenger Talk, video conferencing system Google+ Hangouts, and office spreadsheet app Sheets were unavailable on Monday morning, US West …
I agree on its use as the spousal-conduit. Indeed I believe my wife is responsible for the outage: she tried to get a full history of our conversation (probably to get proof of my wrongness in some regard) and down it went. As we operate a "you touched it last" policy I think the blame is clear (however, she's got nothing to do with the other services - that must be someone else's missus)
All my spreadsheets were fully available to me. They are stored on my domestic local network and automatically backed up twice a day with a 10 day rolling record (on a separate device). I also have a separate copy on an ftp server in a geographically remote location (which I try to remember to update regularly). What is this cloud thing they keep talking about?
It's that thing that is managed by people who don't know you nor care much about you or your needs, but will take your money and promise 24/7 service.
Then they turn around and stop things you need because they scheduled a major upgrade during a work week, which, as any private-company-employed sysadmin knows, is something you do over the weekend so as to not disrupt business.
But they got their business when you gave them your money. Your business ? No worry, Sir, we'll be back online soon. When ? <click>
Google is believed to run as many of its services it can off of a very small set of highly-integrated bits of software. For instance, it has a resource scheduler/cluster fiddler named Borg (new version: Mesos) that I've heard is in charge of the majority of Google jobs worldwide. I've even heard that Larry Page would like the company to ultimately use a single database across all of its services, but that has not happened yet (nor may ever be feasible). That said, we do know that it uses stuff like Spanner and F1 across a huge amount of services. So, when I say monolithic, I mean that the majority of Google's services are believed to draw upon a single mesh of services, so problems in one app may be indicative of a wider problem.
Thanks for the reply.
From the stuff I've heard Google places a lot of emphasis on properly developing and testing its code, which means loosely coupling components. This doesn't in any way preclude the kind of rollout of the plumbing you refer to and associated incidents that integration testing didn't pick up.
Delegating the plumbing (say cache management or replication) for the key applications to standard services is much smarter than reinventing the wheel in each application which I think any developer will have experienced at least once. But this doesn't make it monolithic.
I hate the cloud. I travel a lot and have four homes on two different continents. One of my homes is in a remote location with no Internet service. While traveling - by car, rail, or plane - I am often in places without Internet. It is totally annoying to want to get to mail I have already read but since it is not on my computer it isn't available. The cloud is a bad idea. I have plenty of unused storage on my own system. Why would I want to make it unavailable to me whenever I am not able to connect to the Internet. I venture to say Google drastically lowered its storage price because people aren't migrating to the cloud for the same reasons I don't. Almost everyone I know is not interested in the cloud, doesn't want to give away their data and are terrified of losing it by storing it in some unknown, uncontrollable location. All you silicone valley freaks who live your lives on the Internet think everyone does. Get out in the world a bit.