As a very happy user of PiCloud, I'm sad to see it go. It is a shame they had such A-list VC's (Andreessen Horowitz, Kleiner Perkins Caufield & Byers, and Greylock Partners), who don't have an interest in just a good solid business. They only wanted a huge, home run.
I've been looking around for a good replacement, and there is a gap. PiCloud really had three components. The first was an ability to run code with arbitrary dependencies with no more than a second start up time. PiCloud called them environments. This is really the same problem Docker is addressing. Both were using LXC and AUFS to solve the problem. Unfortunately, Docker-as-a-service is very early days. People like orchardup.com are getting there, but so far only have hourly pricing. There is enginedock.com which has per minute pricing, but is it super early on. I'm in a "Docker gap", where these services are really not at the same level that PiCloud was.
The second was the nice Python API, RESTful interface, and the command line interface for triggering off jobs, querying when the finish, and getting the results. That part will presumably be open sourced, but it is really all tied in to the hosting portion.
Then the third level is to have a bunch of machine you can load balance across, to allow per second pricing on AWS. They had such lovely high memory, high CPU machines to run on. The split up some of the biggest EC2 cluster instances, and you could run each process on your 1/8 share. The nearest current replacement is iron.io's IronWorker. However, they only have 320MB per job, which doesn't work for my application. Also Heroku workers are billed per second, but they only have up to 1GB of memory. I'm hoping that this (new?) entity Multyvac can serve that role successfully. Hopefully, they'll tell us more detail on this soon.
Anyway, best of luck to the PiCloud team. You were great. Especially Aaron who did much of the tech support. Thanks.
Hey Craig, speaking on behalf of the docker project: we are busy filling the "docker gap" you speak about. There are currently many businesses aspiring to offer infrastructure to docker users. What we're doing is giving them a way to integrate directly into docker itself, so you can "docker run" straight to the provider of your choice, either from the docker command-line or remote API. All with minimal risk of lock-in.
You might also be happy to learn that some of the talent behind Picloud is now part of the Docker team itself :)
There are tons of companies that are working on Docker based solutions to my problems. In six months, or a year, there will be plenty of great ways for me to use Docker. The only problem is that PiCloud shuts down on February 25, so I'm in a bit of a hurry. Today, I'm in a gap, and there isn't a great replacement. Specifically I'd need:
* Per second pricing (per minute would be Ok). This makes parallelism "free"
* Large instances with big memory and CPU
* Ideally on AWS (My data is on s3, and it would be good to avoid bandwidth charges).
Nothing seems to fill these requirements yet. Any of you Docker startups out there want to help me out? Any pointers would be great if I'm missing something.
I want to echo your sentiments. We use it for our business and since Aaron told me a few weeks ago that picloud was closing down as a service I've been wondering what to do about it.
The docker gap is one part of it. For me I only really need one environment though (I have a customised one I run on picloud) so that's not the major concern. The bit that wasn't as obvious to solve is moving the python library dependencies from the client to the server. I looked into libraries to handle this but I couldn't find anything obvious. All the other processing libraries use a model where you need to have the code you want to run up on the remote machine before you start.
I recommend checking out CPUSage. They offer a very similar service that's language agnostic. They're pretty early on AFAIK, so they still have a few kinks to work out, but i've found it to be pretty solid.
I think if you read the announcement carefully, it is the standard aqui-hire wording. Dropbox isn't remotely in the same space as PiCloud, and doesn't want to be. They simply "acquired" PiCloud to get their team of developers, which is why they're fine with open sourcing PiCloud. They won't even be using it internally.
the team has decided to join Dropbox...
We’ll be bringing our API-building expertise to the Dropbox Platform, ...
Even though the team is moving on ...
I also hope that Multyvac will be a success. That said, I hope they'll start communicating soon. I can't really judge if they'll be able to be a success... but I hope so.
Oh man, two things: first, I wonder what DropBox will do with PiCloud, and second, I feel like a dodged a bullet. I kinda hate saying that, since it always turns up on these threads, but it's just true.
I had been working seriously on converting a financial application to run on PiCloud. It's mid-November. In the hundred days until they cease operations, I would have had to deal with 10-K filings for my clients and scramble to get the hell out of there, just in time for Q1. What an absolute mess that would have been.
No other time would have been much better, and I know there's an API compatible replacement in the works, but I've been there, and I'm sure something would have freaked out.
Whatever happened to having some kind of reasonable transition period where the buyer continues to support all of its clients? PiCloud, your product was really neat. DropBox, I really wish you wouldn't just hit it and quit it.
I think it is pretty clear from the wording it is an aqi-hire by DropBox. They are just getting a talented team.
I am very close to launching Predictobot.com, which uses PiCloud for all of its computation. I spent last week moving to Heroku after DotCloud's pivot, and now I'll need to replace PiCloud. The web sure is a stable place.
This is an interesting variation on this round's "boomlet". So in the previous dot.com boom there were lots of companies who didn't have revenue so they appeared, until the cash ran out and then fell back to earth. Now we can start a company with much less money, but they go poof in an aquihire when things go squishy.
So was it a pricing issue? or a customer acquisition issue? How is it that so many of these companies get into this squeeze? Is it a chicken -/- egg thing where you can't price it to be self sustaining until you get critical mass? Or is it that Amazon, Google, and Microsoft are just sucking all the oxygen out of this space?
Which brings up another observation: when starting a business, sure take advantage of all those practical and inexpensive tools that can help you build your product in record time. But as soon as you start getting some serious traction and money (income or investment), don't hesitate to replace them with something you have more control over. Hm, it would be interesting to see open-source, self-hosted equivalents of useful services like PiCloud and the like.
At PlotWatt, we've been using the picloud's pickler for a while in our own implementation of the picloud job queue. We don't have all of the bells and whistles that their service did, but running our own queue made it easier to control the details like the number of machines, environment and network configuration, though we tried them out first before their environments feature was set up, so they might have solved most of those problems by now.
Anyway, if anyone is interested in seeing even a bare-bones python function queue based on picloud's pickler, get in touch and maybe we can open source what we have and build it out a little more. Its not open source right now just because its tied into some of our other libraries, but if there is interest, we could probably split it all up.
Sorry to see these guys go, they were a great company to have in the Python space (though they did branch out to other languages, I think).
When we started PythonAnywhere they were right up there on the list of potential competitors, though as it turned out our early adopters steered us more in the direction of hosting and lightweight computation (ping an API periodically and do stuff with the results) rather than heavy-duty computing.
The big problem with VC backing is that they really want $1bn exits. For them, Heroku's $200MM exit to Salesforce was almost a failure. I guess they couldn't see PiCloud going that way, so they chose not to invest again. Better to put the cash into a more likely-looking moonshot.
IMO this is a pity, because there are a bunch of smaller-than-$1bn but still solid companies that are getting starved of funding. The business-to-developer space is particularly prone to that kind of issue; it's less of a winner-takes-all market, so the big wins that VCs want are rarer than they are in other markets.
Yeah - unfortunately there aren't a ton of companies in this space, which is rough. I've started using Domino (http://www.dominoup.com) and it's pretty similar - probably the closest thing out there right now to PiCloud that I've seen. Worth checking out for sure.
I've been looking around for a good replacement, and there is a gap. PiCloud really had three components. The first was an ability to run code with arbitrary dependencies with no more than a second start up time. PiCloud called them environments. This is really the same problem Docker is addressing. Both were using LXC and AUFS to solve the problem. Unfortunately, Docker-as-a-service is very early days. People like orchardup.com are getting there, but so far only have hourly pricing. There is enginedock.com which has per minute pricing, but is it super early on. I'm in a "Docker gap", where these services are really not at the same level that PiCloud was.
The second was the nice Python API, RESTful interface, and the command line interface for triggering off jobs, querying when the finish, and getting the results. That part will presumably be open sourced, but it is really all tied in to the hosting portion.
Then the third level is to have a bunch of machine you can load balance across, to allow per second pricing on AWS. They had such lovely high memory, high CPU machines to run on. The split up some of the biggest EC2 cluster instances, and you could run each process on your 1/8 share. The nearest current replacement is iron.io's IronWorker. However, they only have 320MB per job, which doesn't work for my application. Also Heroku workers are billed per second, but they only have up to 1GB of memory. I'm hoping that this (new?) entity Multyvac can serve that role successfully. Hopefully, they'll tell us more detail on this soon.
Anyway, best of luck to the PiCloud team. You were great. Especially Aaron who did much of the tech support. Thanks.