Dreamhost – How Tolerant Should I Be?
12 months ago I moved Geeknews over to Dreamhost services from Supernova.net. The reason for the move was Supernova’s flakiness in uptime. The last straw in my patience was when their servers went offline for over a week. Not just Geeknews was affected by the outage but so was my family site that was responsible for hosting the family’s e-mail.
Since the move to Dreamhost though it seems as if I’m plagued with bad luck. Numerous times in the last year their file servers have gone offline for over 12 hours at a time, meaning Geeknews along with hundreds or thousands of other sites hosted on that server, or its disk array, have been inaccessible. It also seems that most Friday’s the site goes offline for an hour, probably for maintenance?
A few weeks ago the outage lasted a couple of days, again this Friday evening we were offline until Sunday morning. Each time we’re offline we either get moved to a new server or the existing server is reconstructed and each time the disks are refreshed with a backup image that is 7 days old meaning i’ve lost all files (mostly images) uploaded in that time and need to go back and clean it up…again. So for the 2nd time in two weeks I’m reconstructing the site.
I can’t imagine if I had customers dependent on my site being available like some of the poor people affected by the outage, thank god I just do Geeknews as a labor of love or else I’d be apoplectic right now instead of just very annoyed.
I have a lot of sympathy for the support folks, it’s not their fault HW goes offline, but it would appear the management needs to invest more in fault tolerance, backups, alternative HW in hot standby to switch us over to with just hours of downtime as opposed to days, so they can work on the failed machine at their leisure.
Moving to a new host provider is such a fraking hassle, though I guess I should at least start looking now and staging the content to flip the switch the next time DH is offline.