Does northstar come at the expense of a slight drop in service for standard servers? Options

[Jeff]

Maybe this is the half empty vs. half full balance, but do the Planet's new Alpha dedicated servers represent a slight drop in quality vs the previous generation of Planet Servers? Should there be a "middle class" server class?

http://www.theplanet.com/why-the-planet/ now says the network for non-fully-managed servers is now only covered by a 99.9% SLA vs. the Planet's previous 100% SLA. Isn't that something like 44 minutes of network infrastructure downtime per month now allowed for these servers vs. the 100% or credit of the previous generation Planet servers? That graphic also omits hardware replacement SLA entirely for the Alpha series putting that only on the Northstar side.

If "Alpha" is kind of the value series, I wonder if a new Total Control server breed is called for with fast, guaranteed hardware support, a 100% network SLA, and a 1-hour hardware replacement SLA but focusing on hardware and network perfection only and still leaving normal software management to the customer? Something between Alpha and the fully managed NorthStar (where I can't even find mention prices, which makes me fear"if you have to ask you can't afford it" maybe something aimed at the corporate market but leaving the little guys with slightly less than we had before here)

[JohnB]

I'm glad you asked this Jeff I was thinking the exact same thing.

[Kevin Hazard]

Maybe this is the half empty vs. half full balance, but do the Planet's new Alpha dedicated servers represent a slight drop in quality vs the previous generation of Planet Servers? Should there be a "middle class" server class?

http://www.theplanet.com/why-the-planet/ now says the network for non-fully-managed servers is now only covered by a 99.9% SLA vs. the Planet's previous 100% SLA. Isn't that something like 44 minutes of network infrastructure downtime per month now allowed for these servers vs. the 100% or credit of the previous generation Planet servers? That graphic also omits hardware replacement SLA entirely for the Alpha series putting that only on the Northstar side.

If "Alpha" is kind of the value series, I wonder if a new Total Control server breed is called for with fast, guaranteed hardware support, a 100% network SLA, and a 1-hour hardware replacement SLA but focusing on hardware and network perfection only and still leaving normal software management to the customer? Something between Alpha and the fully managed NorthStar (where I can't even find mention prices, which makes me fear"if you have to ask you can't afford it" maybe something aimed at the corporate market but leaving the little guys with slightly less than we had before here)

That's a good suggestion, Jeff.

The change to the SLA occurred in January of this year following an audit of the legacy SLAs and policies: Power is covered by a 100% SLA and Network is covered by a 99.9% SLA. We were a bit limited by the previous version of our site, but we tried to communicate that distinction as clearly as possible. It would seem that the new design helped out a little there.

The hardware replacement and response time SLA's exclusive to the Northstar division are largely enabled by the design of the managed hosting program ... With dedicated account teams each responsible for a smaller number of customers, those expectations are very reasonable, so we set those expectations externally (while internally, our goals are much more ambitious).

Without a doubt, our internal goal across the board is 100% power/network and hardware replacements and ticket responses completed as quickly as possible, so the SLA is essentially a living and evolving component of our business that enables us to set better expectations as we learn more about our customers and help them build their businesses. The idea of a 100% SLA is thrown around so often in the industry that it has started to lose meaning: instead "we promise (legally-binding) 100% uptime," it starts to mean "we promise to try to provide 100% uptime." I know that doesn't directly answer your question of 99.9% v. 100%, but it's as much background as I have about the decision that was made at the beginning of the year.

As I mentioned at the top, I think your suggestion to have a "middle class" server of sorts is pretty interesting. I know it'll be at least a week or two until we can seriously evaluate what that could look like in the midst of our products and services, but it will definitely be (and already has been) a point of discussion.

[Jeff]

What has always been good about the Planet historically has been that they have stood behind their SLA 100% without hesitation whenever there was a rare problem.

The difference between getting $30 back if there is a 30 minute network outage or router failure and not won't make much different to me.

What I fear though is that the reduction in the SLA may mean that the executives/accountants don't give the datacenter staff the resources they need to maintain the previous service/redundancy levels since not doing so will no longer directly cost them money.

I fear that with the new 99.9% guarantee things may be specified in the future to give at least 99.9% uptime whereas previously 100% was the bar.

Hopefully, this is just a silly and unfounded fear.

[Kevin Hazard]

What has always been good about the Planet historically has been that they have stood behind their SLA 100% without hesitation whenever there was a rare problem.

The difference between getting $30 back if there is a 30 minute network outage or router failure and not won't make much different to me.

What I fear though is that the reduction in the SLA may mean that the executives/accountants don't give the datacenter staff the resources they need to maintain the previous service/redundancy levels since not doing so will no longer directly cost them money.

I fear that with the new 99.9% guarantee things may be specified in the future to give at least 99.9% uptime whereas previously 100% was the bar.

Hopefully, this is just a silly and unfounded fear.

I think it's a completely legitimate fear, but it's one that you don't need with us.

As I mentioned, I wasn't a fly on any of the walls when the decisions were made to standardize the network SLA at 99.9%, but that certainly isn't the "acceptable" level internally (I believe the bar there is currently set at around 105% ). Network uptime is a particularly prickly pear, so I'm going to keep digging a bit more to get a better understanding of why the external expectation was set at 99.9%.

[Tomy Durden]

What has always been good about the Planet historically has been that they have stood behind their SLA 100% without hesitation whenever there was a rare problem.

The difference between getting $30 back if there is a 30 minute network outage or router failure and not won't make much different to me.

What I fear though is that the reduction in the SLA may mean that the executives/accountants don't give the datacenter staff the resources they need to maintain the previous service/redundancy levels since not doing so will no longer directly cost them money.

I fear that with the new 99.9% guarantee things may be specified in the future to give at least 99.9% uptime whereas previously 100% was the bar.

Hopefully, this is just a silly and unfounded fear.

DC Ops will aim for 100% no mater what the SLA states.

I haven't been told about any resource reduction nor see any reason to do so. When we look at ways to reduce costs, we look at improving inefficient and ineffective processes and reducing defects. We don't look at reducing cost by reducing quality. You can read a better description of our internal beliefs at http://blog.theplanet.com/2008/05/02/a-new...-at-the-planet/

One example of a cost reduction and quality improvement:
http://blog.theplanet.com/2008/05/12/new-s...ter-operations/

[Jeff]

Good.

I suppose before I came to the Planet when I had a server at EV1 that didn't have any SLA at all, and the network was still rock solid.

And I suppose one reason for reducing the SLA might be that in the event of an unavoidable 30-minute network outage for example, the credit would be so minimal even under the old 100% SLA for the short outage that it wouldn't really increase customer satisfaction so thus it would be just throwing money away.

[Brooke-Sales]

The change to the SLA occurred in January of this year following an audit of the legacy SLAs and policies: Power is covered by a 100% SLA and Network is covered by a 99.9% SLA.

The 99.9% network uptime guarantee has been the SLA for our Built-in Support servers since before January of this year. It was 99.9% at least as of the launch on January 15, 2007, though I think before that too. In fact our "Economy" line of servers had no SLA (nor did any EV1 servers pre-merger).

There was, however a 100% network uptime guarantee SLA included with our Enhanced Support package, so the only thing that has changed in regards to the SLA due to this relaunch is that with no more Enhanced Support option you cannot upgrade to a support package that includes a 100% network uptime guarantee SLA. If, however, you signed up for Enhanced Support prior to Tuesday we'll still honor the SLA you were given.