Server Downtime Earlier

My website was down beginning last evening till this afternoon. Dreamhost apparently had some DNS issues which took forever to resolve. Outlook kept complaining that it couldn’t connect to my email server. When the DNS issues were finally starting to resolve, I started getting my emails in chunks as email servers retried sending them to me. I also got back a bunch of unable to send emails (people spamming using my domain). I’m guessing these emails were also trying to reach other Dreamhost customers.

Everything’s back up now, but I’m starting to get tired for all this Dreamhost nonsense. I’ve decided that after my contract is up, I’m probably going to switch over to 1&1 and purchase a business plan. Space and bandwidth might not be as much, but I’m only using 8GB out of 256GB on Dreamhost. I use about 65GB/month, but my plan comes with over 3,000GB. For the 1&1 Business plan that I plan to upgrade to, for about the same $10/mo plan, I still get 250GB of webspace and 2,500GB of bandwidth.

Plus the fact that I haven’t seen 1&1 down for half a day has really given me confidence during the past 3 years that I’ve used them.

You know something’s gone awry when DreamHost has to offer a new add-on for $15 more per month so that you can get guaranteed CPU and memory. I got an email awhile back introducing DreamHost Private Servers. For $23/month (assuming you’re on the 2 year contract for $7.95/mo), you get a guaranteed 150MHz CPU and 150MB of RAM.

For that amount, I can get a 1&1 Developer Hosting package.

You know, DreamHost is the first webhosting service I’ve hit into CPU/memory issues and it hasn’t always been like this. My gallery and blog used to coexist fine on the same account. Now I have to split them apart because combined, they’re consuming too much resources for 1 account.

Many people have said they oversell their web hosting, but I’ve always thought, who doesn’t? But I’ve never hit a problem until a few months ago, that I was getting 500 Internal Server Errors because they have a CPU/memory monitor program that kills anything that exceeds a certain CPU usage during peak hours. One person even calculated if he was to serve a static page 24/7 at the sustainable speed and below the CPU/memory requirements, he’d never be able to reach the allotment of bandwidth that he has purchased. I can’t find that blog entry anymore, but I did find this: Dreamhost Sucks At Hosting

3 Replies to “Server Downtime Earlier”

  1. I’m on DreamHost too, of course — same as you, I’m a bit fed up with them. The CPU cycles limit is a big part of my frustration, but so is the downtime (my own measurements put the uptime for my server around 96.5% — DH doesn’t dispute this). BTW, that same site you linked about DreamHost sucking doesn’t have very kind words for 1&1 either.

    Anyway, when my DH package/contract/whatever expires in October I’m thinking of moving back to HostPC. I used em for two years and only moved because their best package didn’t offer enough bandwidth. But since then they’ve expanded things. They’re not the cheapest, but they have a way better track record than DH IME.

  2. Looks like that was the case. Akismet did catch it as spam.

    I’ve had 1&1 almost 4 years already and never really hit into CPU problems. I mean DreamHost was pretty good until I started getting 500 Internal Server Errors. If DreamHost can maintain the quality of not showing me that error, I’d be happy.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.