Previous Hosting Problems

Suffering from coders’ block and feeling a little down(probably an oncoming cold), i thought I’d write a little.

These past few days, there has been no less than 4 people who came to our house all sick and germ-y. And suddenly i turned into Mr. popular where everybody wants to talk to me and hug me and kiss me…yes, even people i don’t know seem to want to do that. I mean that’s really bad. When i get sick, i make sure i don’t go anywhere or see anybody until i recuperate so that i wouldn’t get anyone else sick. I can just hear my immunity system right now saying “TO WARRRRRRR!!!!”

So going back to the technical stuff.

Every time i read about python related sites, i also read the term “Webfaction”. And after a recent issue with my previous host “Rails Playground”, i decided to move to “Webfaction” and see what the fuss is about.

“Rails Playground” is not a bad host, but there was an incident where some of my data was lost and the way that they handled the issue didn’t really scream professional. The server that i was one was unresponsive for over 24 hours which isn’t really a big deal for me. After investigation, it turned out that the entire RAID array was corrupt. I think that the probability for that array configuration to become corrupt is something like 10% or so. Yet it did go corrupt. problem…it happens. Obviously in a high stress environment such as a data center, the probability of hardware failure will rise significantly(i even read somewhere that google has to maintain about 10% of their 100,000+ computers each month)

So the “Rails Playground” people submitted a replace request and the array was replaced promptly. Only when trying to get back the customer’s data, they couldn’t. I am sure that they wouldn’t be dumb enough to store the backups on the same array, would they?

After i submitted a support ticket for my unresponsive website, they said something like “yeah, the server is down, please cooperate with us” and for the ticket about the data(which at the time was only a blog post but it is usually more), it went something like “yeah, we couldn’t read the backup data so you lost two days or so of data.” They weren’t lying mind you, but they were saying it like it was a very normal and usual thing. So after this incident, i decided to move hosts again having only been with “Rails Playground” for two months.

For my website i can usually tolerate such mistakes since i know that there is no guarantee for hardware stability and software security. But i expect that paying for a managed service really means that it is a managed service. Meaning the server is secured as much as possible, hardware is monitored and network is observed. And when a mistake does happen, and they also do. Mistakes are fixed and explained and something somewhere altered so that they don’t happen again.

By reputation, “Webfaction” seems great. And their support also seems friendly so far. Although they don’t offer a lot of resources, for the price it seems alright, and i will write back if i had any issue once i run some python apps on it. On thing to note is that everything seems faster after moving to their servers. My website, ssh access and even the mail but maybe that is just a placebo effect.

Like i said before, i have a few ideas and a few projects that i want share and i will execute them as the time allows. And after my wordpress problems(which should be fixed now), i was thinking of running on an all-python stack. But i am still considering my options short of writing my own CMS and blogging platform in python. So if anybody has any suggestions, contact me.

More later….