Slashdot Mirror
Google Apps Gets a 99.9% Guarantee
David Gerard passes along a posting on Google's official blog announcing that they have extended the three-nines SLA for the Premier Edition of Google Apps from Gmail alone to also cover the Calendar, Docs, Sites, and Google Talk services. 99.9% uptime translates to 45 minutes a month of downtime, and the blog post puts this in context with Gmail's historical reliability, which has been between three and four times as good over the last year (10-15 min./mo.). It also claims, based on research by an outside group, that Gmail's historical reliability beats that of in-house hosted solutions such as Groupwise and Exchange, on average. Reader Ian Lamont adds an article in The Standard that digs down into the details of the SLA, revealing for instance that outages of less than 10 minutes aren't counted against the monthly 45 minutes.
Most likely it's the time for node crash detection and load balancing to take effect.
If service is that bad or intermittent, nobody would buy service there.
It's called a cluster, "The cloud" is a really annoying buzzword for software as a service.
/Mikael
Greylisting is to SMTP as NAT is to IPv4
Yes, but what is the average company's internet downtime verses their LAN downtime for a single-campus outfit?
So instead of LAN / Exchange Server (or whatever is being used) you now have LAN / WAN / Google downtime. WAN gateway downtime is probably the weakest link in the chain, so wouldn't the total downtime be greater using something internet based?
Better known as 318230.
Google is a company. Saying "Google doesn't have 100% uptime" makes as much sense as saying "Microsoft takes 40 minutes to install". What specifically are you trying to say?
The 99.9% guarantee is great, if there's someone to talk to who'll actually look at the problem when those three 9s aren't met. Otherwise it's marketing propaganda.
These posts express my own personal views, not those of my employer
My big concern with this type of offering is it increases a companies dependence on their internet line. If your network is down not only can't retrieve files, email or browse, you now can't work on productivity software either. Essentially if your doing a job that requires a computer in this environment you can't work whenever the internet or network has a hickup. I like having something else to do in the rare instances where the network isn't working right.
Add to that the fact that wireless/laptops are becoming of larger importance in companies (and wireless is flaky at the best of times IMHO) you're really courting disaster not just in terms of outages but in terms of accidental data loss. Say your not so gifted technologically colleague decides to walk over to your desk with their laptop to show you the spreadsheet they've been working on. They get out of range of the router that they were using and presto session time out and the chance of data loss.
As a commercial user of Google Apps, I have observed this not being the case. GMail does go down, and the cause is not our connectivity. What's worse is when there is a problem, all the 'phone support' does is tell you to post on their forums... not impressed.
was their claim that this is 4x less outages than on-site-maintained Exchange or GroupWise.
(Notes, of course, gets 45 minutes of uptime a year.)
http://rocknerd.co.uk
I achieved four nines (%99.99) 8 years ago with Netscape's broken mail server "Suite Spot" running on a (at the time) three year old Sun E450 with 4 gigs of RAM. As I recall, it served about 120,000 clients on a large cable network in Chicago.
This whole "new web" thing is very pretty, but it seems like about three steps back to me.
There'd be no need for a Beowulf-type cluster in this case.
Have a bunch of machines running identical instances of Apache, and randomly fire requests at them individually. This balances the load, and ensures that the servers themselves aren't a single point of failure.
It's quite a bit more complicated than this in reality, although you should get the basic idea.
Beowulf is typically used for clusters that seek to emulate a supercomputer (usually for scientific number-crunching), rather than a server. For this reason, something like Google's setup would more typically be referred to as a "server farm"
-- If you try to fail and succeed, which have you done? - Uli's moose