Don’t Quibble with Quimble

Very nice email from Arman Zakaryan. MediaTemple response and my reply is below.

This is a follow-up message regarding this morning’s availability issues with your Application Container. We very sincerely apologize for the extended downtime that you experienced. Please notice the $20 credit I have just placed on your account as a token of apology.

That being said, we understand that you just want to run a business and do not want to deal with issues such as this. We are doing our absolute best to ensure a smoothly-operating hosting experience for not just you, but all of our clients. I work with our staff every single day (even on my “off” time) and see the amount of hard work that goes into our operations. If our level of availability is causing a serious problem with your business, we understand if you’d like to try out other hosting solutions. Your business with us as a client is extremely valuable and appreciated, but we understand that you must do what’s needed to ensure the success of your own venture.

It was a pleasure to be of service to you. Thank you for using (mt) Media Temple.

Best Regards,

Arman Zakaryan
Senior Technical Support
(mt) Media Temple
Your response: 2007/06/12 18:20

My response:

I truly appreciate this response. I get frustrated, and I’m sure I make you frustrated as well. My biggest gripe from you guys has been the lack of communication regarding long down times. Every host goes through a growing pain and you guys are doing something new. I understand it can be hard and that there are unforseen problems. However, most of what I get are form letters written in a way that seem like they cover-up serious underlying issues. The gridserver is supposed to scale like crazy, but if that’s possible then it should be possible to have 99.9% uptime. If there is serious over-capacity, then you should be able to switch things to other servers.

Sorry… that’s me getting frustrated again and straying from the point. My point is: I want to know when I should expect longer-than-normal downtimes. I want to have a back-up plan provided by you for when things go wrong. I want a down-page that says “oops the gridserver fucked up again… we’ll be back shortly.” I want an email that explains “this F5 load balancer is hard to configure, and we fucked up” or “Doees anybody have any ideas regarding xen virtual machines configured this way?” You have a wealth of customers that know what the hell they’re doing… use us.

It’s VERY frustrating when you’re public face uses things like “brief down-time” for 12 hours. I think from the email above you would hate that as well.

Talk to me. Talk to your users, talk to the public. We’re very forgiving if there’s enough information being passed around.

Thanks,

Topper