Sad that I CANNOT RECOMMEND Dreamhost

tl;dr summary: My small personal site on DH was very frequently down. When it was up, it was dog slow. I moved to a new host, pay the same price as before, and I’m enjoying 100% uptime and a 2x faster site.

I was a longtime Dreamhost user on and off over the years. There’s a lot to love:

  • Lots of features and storage / bandwidth for a low cost.
  • Free support from usually friendly service folks.
  • A control panel with tons of options for beginner and advanced hoster alike.

But as I discovered to my sadness, Dreamhost is like a restaurant that has attractive decor and friendly waiters, but either fails to actually serve your food… or brings it a few hours later than you ordered.

Over my last months with Dreamhost, my small personal WordPress site (<50 visits a day) was down every week, often for hours (I know this thanks to external monitoring and personal experience). Even when it was up, the site was shockingly slow; simple operations took seconds (instead of fractions of a second) in the WP admin panel, and site latency for my visitors was also unacceptable.

But I persevered. I kept filing respectful tickets or having chats with Support, and they kept telling me:

  • Hmm, must be one of your plugins!
  • In either case, we’ll just reboot the apache instance.

After many instances of this, they moved me to a new apache instance, and the site stayed up for… a full 24 hours.

Like an abusive relationship, I was beginning to worry that indeed something was my fault. I’m a pretty geeky tech guy myself (have been online for more than a quarter century… yes, pre-public-access Internet days), but I asked a friend to take a look at my WP site to see if there were any red flags. There were none.

I found another host that I will not name here that was highly recommended. They kindly moved everything over for me, and then my eyes were opened: everything was faster… more than twice as fast (as measured myself, and via external testing tools)! And in the weeks I’ve been on this new host, there’s not been a single minute of downtime on my site.

Oh, and I’m paying exactly what I paid DH.

So, sorry prospective DH customer, I cannot recommend this hosting company. They seem like they’re a well-meaning, hard-working bunch, but obviously something has fallen through the cracks. And that something is reliable servers.

That’s a fair enough verdict. Though it also worth noting that every company from Amazon to Zoopla will have some customers disappointed in their product. (And by virtue of the fact they are in business, numerous more customers who are happy.)

Yep, and that’s yet another reason why I’m sad about leaving Dreamhost. Clearly they CAN do right, and ARE doing right by a lot of people. But boy are the ‘edge’ cases painful ;).

I agree with your nostalgia and identify with the same problem (which has been on and off, too often on, for the past 18 months). Always told it’s me and traffic etc. and that i should use VPS but it makes no sense when some of the sites get less than 30 page views daily.

At this point, I hate dreamhost and would not recommend it to anyone. I am having slowness problem after problem after problem. At one point I did a trial of the VPS and it made no difference. How am I supposed to live with a Wordpress backend that takes over a minute to save a page, or it times out trying to save a page? I have very little traffic, and I rarely have more than one user browsing my site at one time. I use it the most of anyone! I just don’t get it. I have done several of the things they said I should do to reduce memory usage in my site, including moving two of my subdomains to separate users, and removing a couple of plugins whose functionality I was able to replace by coding it up myself (this work is still continuing). And there is no change in performance whatsoever. I get one day where maybe, my site loads at a normal speed, and then it is back to being so slow it is unusable. I am now in the process of asking to be moved to a different server, but customer support is hesitant to move me for fear I will have the same problem on a different server. Are you kidding me? Is this what I am paying for? I also had several chats with support who promptly told me it was my plugins, and I too kept thinking it was me. I am so fed up with dreamhost all I want to do is scream. Dreamhost is supposed to be Wordpress friendly and be one of the best places to host a wordpress site. But that has not been my experience at all.

ThatAdamGuy, is there any way you can email me the name of the company that you moved to?

So what you do is disable all plugins and change the theme back to Twenty fourteen. Does the problem persist? If yes, then lean on support for that move to a new server, if no… then figure out how to optimize. This isn’t an easy answer either sometimes plugins A, B and C are fine but if D is also running it mucks things up… likewise with the theme.

I haven’t hear many complaint related in the last 30 days so I don’t know if they have made progress on it… but there’s also this apache glitch that wrecked havoc for over a year. Are the delays you experience in multiples of 30 seconds? Meaning 30 seconds, 60, 90… etc?

Hi LakeRat, thank you for your reply. Last week I did a lot of searching on these forums to hopefully find some sort of answer or insight and I did run across the posts on the apache glitch you mentioned, and I want to implement some of the monitoring code that was suggested. But with competing in horse shows on the weekends and with the other website updates that my user’s want, and with holding down a full-time job, I haven’t had time to work on it yet. But I really want to try something like that that will help me understand what is going on. I have also not had time to turn off all of my plugins and turn them back on one by one and try re-trying to edit something on the backend each time. That exercise is extremely time consuming as well. But, it is still on my list of things to try. However, I have the same problem with another very small site that uses the same theme and has only 1 plugin, so it is somewhat lower on my list of things to do.

It can take time to maintain a site, of course it takes less time if everything works right.

The first steps you need to take are just rip out your plugins and theme if the problem goes away then the problem is in your combination of theme and plugins. If it stays around, the fastest way to get an idea if the apache thread is involved is to simply time the delay… is it a multiple of 30?

Well, the delay is generally longer than 30 seconds, so I don’t think it is a multiple of 30. But, without instrumenting my code to report some timing information I can’t say for sure. I hope to have some time to work on this this weekend. It’s just been a tough week.

Unfortunately, I am also dealing with an issue that I created for myself while attempting to do what DH support told me to do. And that is, to assign only one user to each subdomain. I did that part ok, and moved my development subdomain to a different username. And since my development site and production site could no longer share files easily, I used this as an opportunity to switch from git to svn (because I am more comfortable with svn than git) so I could share files through an svn repository. Only now I found out the hard way that you can’t use the Wordpress automatic update feature to update your plugins if you have them in your svn repository (which I do because I have had to modify the code of some of them to make them able to handle custom post types, or add a new filter, or something simple that I really need). So then I had to spend time restoring my working copy of my respository, and searching online for ways to prevent the .svn directories from being deleted when you update the plugins from the wordpress admin, found two solutions, tried them both, neither worked. One was an old wordpress plugin called svn-auto-update that doesn’t work, and the other was a hack to the wordpress core (that I found on that didn’t work either. And so far, I haven’t found anything else out there. I could have saved myself a lot of time if I had just updated the plugins manually. But no, stupid me, I had to try to be clever and find some code to do it for me. So tonight I will just update them manually and get back to making progress on my site. I hope to have time this weekend to look at the slowness issues again and try to get some timing information.

Thanks again for writing …

As a customer of Dreamhost for 12 years, I sadly have to agree. The quality just isn’t what it used to be. Now, if I can figure out how to run two versions of PHP on my new VPS I’ll likely be moving everything off. I can’t tolerate websites running/loading this slowly, and the fact that they either have no monitoring tools to let them know when servers are spiking with runaway processes, or as I’m beginning to suspect, they let it happen to push people to VPS solutions.

Just to note, it is still causing havoc. Two websites on Dreamhost which I abandoned (as unusable) have both been going up and down like yo-yos over the last few months.

One was slow (multiple of 30 secs response time) for most of April, and spontaneously cleared on May 3rd at approx 04:35 UTC (and has stayed fast since then). The other has generally been doing a little better, going slow only for a couple of days at a time and then spontaneously recovering. It most recently went slow yesterday, May 13th approx 19:20 UTC (and is still slow as of now).

However, the load average (both sites are on the save server) has been becoming increasingly awesome over the last few weeks. It tended to range between 10 and 20 through April but then started reducing. In early May it was in the range 5 to 10, and for the last couple of days has been in the range 1 to 5. I’m guessing this means that DH has been progressively moving people off this server (pirates) as part of their move-server project, and presumably will move me soon.

Will be interesting to see what happens on the new server.

I got moved last Friday night at 11PM. I would have expected a move to occur more during west coast business hours, than 11 PM on a Friday night… but whatever. The email announcing I would be moved arrived bright and early on a Monday morning 2 weeks ago.

I haven’t had much time to investigate yet, but on the old server pingdom reported 2 outages in 30 days prior to the move, totaling about 10 minutes. During the roughly 108 hours since the move pingdom has reported 10 outages totaling 2hr 14min.

I haven’t checked much yet, including deciding if the issue pingdom is reporting involves the apache glitch. I did notice that all my sites ended up in the same apache instance this time, that’s different than it’s been with previous server moves. I wonder how many more sites share the new IP and what they are, tools like take a bit to update after a shuffling like that due to the way they gather the database.