Comments on: Google Gaffe: Gmail Outage Shows Pitfalls of Online Services

By: Putting All of Our E-Mail Eggs in One Basket: Gmail Down Once Again » Basex Blog »

Thu, 03 Sep 2009 00:29:16 +0000

[...] The last major Gmail outage was in February, unless I missed one since then. This week’s would be the seventh major outage in one year. Most of what my colleague David Goldes had to say then still holds, so in the interest of brevity, I’ll ask you to continue here. [...]

By: IndependentProfitCenter

IndependentProfitCenter — Fri, 01 May 2009 13:31:21 +0000

Hello. Great job.

By: IndependentProfitCenter

IndependentProfitCenter — Fri, 01 May 2009 13:04:47 +0000

Thanks for this

By: jidanni

jidanni — Mon, 02 Mar 2009 22:42:29 +0000

No problem. If it breaks, just get their tech on the phone. I’m sure
the number is listed somewhere on there on google.corn.

But seriously, even with best intentions, there is little chance for
the average user to get a word in or out edgewise. Scales and
economics just don’t permit it.

So it’s just “luck of the sheep herd”.

By: Tobin Maginnis

Tobin Maginnis — Fri, 27 Feb 2009 17:43:42 +0000

Sorry, but this appears to be a silly expectation. Google has an
elaborate EULA that makes it abundantly clear that use of Gmail is
*not* designed for business/corporate data integrity much less
corporate-level security. Nevertheless, corporate uses flock to Gmail
even forwarding their corporate email to a Gmail account and, using
Gmail’s “reply as” feature, providing a thinly veiled “reply” from the
corporate account.

The simple truth is that there is no contract with Google, implied or
otherwise, and you get what you pay for.

As for a “pitfall of Cloud Computing,” the reliability is based upon
system design tradeoffs and, again, you get what you pay for.

By: Ramon Galofré

Ramon Galofré — Fri, 27 Feb 2009 11:55:41 +0000

let me say that this sentence is a nonsense:
“Unless cooking is your core competency, there is no reason to keep that operation in house. But e-mail is the lifeblood of almost every organization today”
For almost all companies e-mail is not their CORE COMPETENCY. Therefore, why email cannot be outsorced? Place the right SLAs with your vendor.
It does not matter the size of your business, you will have more downtimes and headaches keeping your email in house that outsourcing it to a company whose core competency is email.
In most cases the decision is just strategic or based on a luck of confidence.

Ramon Galofré

By: Michael Winser

Michael Winser — Fri, 27 Feb 2009 11:17:25 +0000

Re Kenneth Hoffman’s comment:
It’s different because:
1- It’s Google
2- Because Google doesn’t just want to be your mail provider, they want to be everyone’s mail provider. When you provider has downtime you notice. When Google (or any other massive provder) has downtime everyone notices
3- Massively scalable cloud services aren’t supposed to fail the way “normal” but distributed systems do.

By: Serge Egelman

Serge Egelman — Fri, 27 Feb 2009 11:06:45 +0000

The existence of “Google for domains” seems to undermine the argument
that Gmail “is not designed for business/corporate” use.

http://www.google.com/apps/intl/en/business/index.html

Given that they are charging businesses to transfer their inbound
email to Google…

By: Kenneth Hoffman

Kenneth Hoffman — Fri, 27 Feb 2009 07:04:22 +0000

I am not defending Gmail, but how is this different from outsourcing your email to another company? For example, I am a one person tax and accounting firm. I am more than capable of running a mail server in house (EE in a former life), but I outsource to mailtrust.com. At times they had issues, but the difference is they advised their subscribers of the problem and kept us updated.

I do have a Google app account, but I have not heard word one from them on the most recent outage.

In you telephone analogy, I am current having issues with AT&T. They are instilling fibre in my neighborhood. I work from my home. I do have a business account and a SLA. All was fine until 2:30 this afternoon. Internet is down. I am told it will be 3-5 days before I *may* get my connection back.

I walked down to where the telco box was being *upgraded*. One of the techs said the phone switch over went flawlessly, but the DSL circuit switch over had an “anomaly”. For now I cannot work from home. My daughter cannot submit her papers to her university, she could not attend her online class tonight from home and my high school aged daughter could not do her research for a paper. Fortunately, my daughter could log into her class from the university library, and other daughter was also able to conduct her research from the library.

Perhaps its time I consider a redundant cable connection.

By: Brent Frei

Brent Frei — Thu, 26 Feb 2009 22:48:23 +0000

Relative to “everyone going down at once”, the impact could potentially be felt downstream by the third party applications published through Salesforce and Google. With their app development platforms, they are carrying a few thousand vendors with them. These apps are frequently reliant on Salesforce and Google’s core services as integral to their apps. Most are not solely reliant on AppExchange or App Engine as their delivery vehicles, but an increasing many are.

The timing of your post coincided with one of my own relative to the typical SaaS company decisions process around hosting through one of these giants.

http://www.smartsheet.com/blog/brent-frei/are-salesforce%27s-forcecom-or-google%27s-app-engine-good-distribution-vehicle

When it’s all said and done, a few hours of down time a year, are almost always a viable trade-off for the distribution and development advantages gained.