Joyent's blog covers its cloud platform, SmartOS, Manta object storage, and Node.js production experience.
HN commenters mention Joyent mainly for SmartOS, ZFS integration, KVM support, and Manta storage, describing them as stable and inexpensive.
Joyent Accelerators ( http://www.joyent.com ) are flexible, and they're the "official" Ruby on Rails host. They also happen to offer the best control panel (Virtualmin, of course). The only negative might be that Accelerators are Solaris systems--I don't have any complaints about that, but if you're accustomed to Linux, it might be a bit of a learning curve.
Slicehost ( http://www.slicehost.com ) has a good reputation. Their Xen systems are not over-subscribed and they know their business.
Media Temple (http://www.mediatemple.com ) had a good reputation until recently...but I'm sure it was a temporary setback. They're technically savvy. They're using a kind of quirky virtualization layer (Virtuozzo) which, when over-subscribed even a little bit (and sometimes even when not), exhibits serious memory issues.
Gridzones ( http://www.gridzones.com ) is another Solaris zones based host, and they have some reasonable plans. It's a smaller shop, so probably gives more personal service. Tell Phil that Joe from Virtualmin sent you (I dunno if it'll get you a discount, but he'll be happy to hear from you). They start at $15, which is hard to beat (you probably want more space than that, though...). They're also Virtualmin friendly, when appropriate (i.e. when you have enough space/memory for full service hosting).
If you're wanting to spend a little more, The Planet and Server Beach both offer dedicated boxes starting at around $69, I think. I've used both for years with…
It seems to me that the workloads on Digital Ocean are not fully utilizing the block device. FIO shows about 50% util for DO, and 81% for Linode on the queue depth=1 test, and 70%/82%, respectively, on the queue depth=8 test. Which seems to indicate a bottle neck somewhere else (at least on DO), and that seems to align with OP's statement at the end, where it seems DO is being capped. Also as it was pointed out, the sizes of these VPS are not even mentioned, and most likely the working set size of 128MB, might be either too small or too big for any of these VPS, making even harder to see the value of these results.
Brendan Gregg has excellent info on these topics,
http://www.joyent.com/blog/benchmarking-the-cloud
http://dtrace.org/blogs/brendan/2011/05/11/file-system-latency-part-1/
http://dtrace.org/blogs/brendan/2012/10/23/active-benchmarking/
This is great in so many ways:
- New logo! Much better "brand" than previous, and look good in small sizes, although I will miss the cute visual pun of GIT letters being refresh, addition and subtraction symbols. Although, both favicon and logo image would look better without half-transparent-pixels in vertical branch.
- New font! Adelle from Typekit, not from Google Fonts. They took it seriously, and it's a fantastic choice.
- The whole branding and theme and everything is way better than what was before, and arguably the best looking site for any DVCS. (maybe took some design inspiration from http://www.joyent.com/products/ ?)
- Built on Rails, and open sourced: https://github.com/github/gitscm-next That's great, and hopefully will allow the community to contribute more to the site itself.
I love the idea of fund-raising from customers. There are a few different angles on it, and one memorable one was the TextDrive thing[1][2].
That turned out a little bit controversial later[3], but it illustrates some nice, creative thinking. Is there a way you can get your customer to commit some money today, to an extent that they might not otherwise, based on future considerations? If you can do it without creating a scenario that bankrupts you in the future, it might be worth a shot.
[1]: http://www.joyent.com/blog/celebrating-textdrives-three-years
[2]: http://photodude.com/2004/06/01/textdrive-or-how-to-raise-40000-in-4-days
[3]: http://techcrunch.com/2012/08/16/cloud-computing-company-joyent-leaves-early-supporters-out-in-the-cold/
Funnily, Bryan's the same person who also showed contempt in crucifying one node's largest OSS contributors in public is now seeking to attract more contributors to the same project?
The 2 character pull-request that took down libuv's largest and one of node's largest ever contributors: https://github.com/joyent/libuv/pull/1015
Bryan's public Crucifixion of Ben: http://www.joyent.com/blog/the-power-of-a-pronoun
Ben Noordhuis's Response: https://github.com/joyent/libuv/pull/1015#issuecomment-29568172
and HN Story of his Departure: https://news.ycombinator.com/item?id=6845286
No thanks, I can't see myself ever touching or recommending Joyent's services as a result.
Disclaimer: I'm the CTO at Joyent.
For whatever it's worth, we completely agree with the sentiment (and I like your "blue collar apps" term) -- and we deliberately have designed Triton[1] for ease of use by virtualizing the notion of a Docker host. I think that the direction you are pointing to (namely, ease of management for very small deployments) is one that the industry needs to pay close attention to; the history of technology is littered with the corpses of overcomplicated systems that failed because they could not scale down to simpler use cases!
[1] https://www.joyent.com/blog/triton-docker-and-the-best-of-all-worlds