11 ms·

Compose Is Joining IBM

172 points by thomcrowe 11y ago

ukd1 11y ago
Congrats guys!
- thomcrowe 11y ago
  Pretty exciting!
nfriedly 11y ago
As an IBM'er, I think this is awesome news! Cloudant is awesome for CouchDB, but if you want mongo or redis, we didn't have a lot to offer before today. Looking forward to working with you guys!
compostor42 11y ago
I've very curious what the general "impression" regarding IBM is around HN? They are not discussed much around here.
- ahoff 11y ago
  IBM is huge, and it's hard to give a general impression around the whole company that makes any sense. I can only really speak about CDS (the division Compose is joining and Cloudant is a part of) but I can say it's a really exciting place to work. CDS is about as fast-moving as you can get in a company this large, and the leadership really understands the shift to SaaS. We have a great team and a lot of autonomy to make decisions outside of the general corporate superstructure. I'm biased but think it speaks highly that Cloudant people (Adam Kocoloski and Derek Schoettle) were tapped to lead this group.
  - mbroberg 11y ago
    It's a huge place, but I agree with ahoff, having come over along with the Cloudant acquisition. I feel pretty good about our little corner of IBM in CDS (cloud data services). It's not a bad place to end up if you're at a data management startup.
- tdicola 11y ago
  IBM is a financial engineering firm, not a technology company anymore: http://dealbook.nytimes.com/2014/10/20/the-truth-hidden-by-ibms-buybacks/?_r=0 http://dealbook.nytimes.com/2014/10/20/the-truth-hidden-by-i...
  - ahoff 11y ago
    If you only care about stock price, and even then no. IBM is a large company that does a lot. They put more money and people towards pure tech than most firms, especially in places like IBM Research.
  - devy 11y ago
    What people remember them the most, is their R&D. The beat Intel to create the first 7nm chip process recently.
    - derwiki 11y ago
      The Almaden Research Center is still a top-notch research facility.
    - timelined 11y ago
      Between Alamaden (San Jose) and Yorktown Heights (NY) in the US, IBM is one of the last few industrial basic sciences/engineering labs in the US, resembling Bell Labs and Xerox PARC from a few decades ago.
- rdtsc 11y ago
  I think IBM is turning around. It missed the cloud shift by quite a few years, but they know it, and they are re-inventing themselves, which is quite commendable for a large 100+ year old company. HN often lives in the startup + Facebook + Google + latest tech trend bubble. But there is a whole world of enterprise + government (DoD) tech world out there that exists, but it is not talked much about here, unless it is being made fun of perhaps.
  - abutler 11y ago
    IBM is no longer just focused on enterprise and government. There's a big focus now on startups. Take a look at the Bluemix Garage where there's IBM engineers embedded in Galvanize in SF to work with startups but also help enterprises bridge the gap and learn from them; and also look at digital.nyc and tech.london. Both are IBM initiatives. So... IBM today isn't like the IBM that most people imagine and I am sure that the IBM of tomorrow will be completely different. You don't get to be over 100 years old and a tech company unless you can reinvent yourself to stay relevant.
    - rdtsc 11y ago
      That is what I was trying to say, I was just responding to what I think HN crowd's perception is of IBM, which I think is not accurate. I think it is really great to see a company of their size and age change course and adapt.
- roymurdock 11y ago
  My impression is that they have almost fully divested from hardware and are doing whatever they can to boost their consulting, services, cloud & analytics businesses - hence this latest acquisition, on top of many other cloud purchases. This is smart, given that hardware is a commodity with low and falling margins, whereas it's very hard to price heterogeneous services such as consulting, integration, etc. They've done some shady things to cut costs, but they needed to pivot pretty drastically in order to maintain their position and restart a growth cycle. Ultimately, I think the "big data" market is overhyped, and that IBM is playing into this hype. People are gobbling up data storage and analytics services left and right, but are the valuations equivalent to the value added to a company by storing and sorting a ton of (potentially useless) data?
- jdcskillet 11y ago
  I just migrated from working directly on the IBM Mainframe in COBOL (gasp) to my true love of .Net and JS (yep, both of them) within the same company. I have been on a number of greenfield projects with IBM and quite frankly, my opinion of them differs based on the area of the company you are dealing with. For instance, I love dealing with the CI guys at Urbancode Deploy, whereas getting similar information from their Rational counterparts (Rational Team Concert) was like pulling teeth. Pretty much every experience I had with the Rational teams was terrible because they would hide information to sell more consulting services or additional products (even though they guaranteed the feature was available on our existing license). I despise the documentation sifting and information gathering required to use IBM products (developer works is a terrible site). However, I have much respect for the engineering that goes into those products and it is a joy when you get to work directly with their engineers to solve a problem. Good luck compose, I hope you get to be autonomous like Urbancode and not stuck in the sand of the Rational teams.
  - davidgerard 11y ago
    Rational's pricing model has always been "how much have you got?"
- teddyh 11y ago
  “That company is large because it is large. […] By not seeking to strive, it conquers without effort.” – The Tao of Programming, chapter 8, Geoffrey James, 1987
- fennecfoxen 11y ago
  My impression of IBM as a place to work is that it's hopelessly bureaucratic in a way that no simple reform is going to be able to fix. For instance. Once upon a time I had an internship there, working with a particular product dealing with server management. As part of the setup process required to customize servers, users were directed to go to a random university user's home directory to download one particular older version of some open-source boot-utility software (newer versions wouldn't work). This is, of course, a terrifying distribution mechanism. Actually hosting and distributing this file wasn't something that IBM people were willing to do; you'd need to run it through the Lawyers. (This probably feeds a Not Invented Here syndrome.) That said! At least the internship was with the Software division. The way I heard it, Services was 90% of the revenue and Software was 90% of the profit. Things are a lot better in Software. Still a damnable maze of cubicle farms, though.
- chris_wot 11y ago
  IBM are increasingly pissing off their customers since they retrenched a good portion of their experienced workforce. I have a friend who is a CIO of a large company in Australia and he is increasingly getting annoyed with IBM's SAP support and is hunting around for alternatives. He probably won't move from IBM completely, but they'll be getting less and less business from him. I mean, the other day their entire datacentre's network link was flapping, for hours and hours. It wasn't just his company that was affected, it was affecting even larger companies. IBM's ability to deliver is going down the plughole. They've always been very expensive, but have been able to justify it because they were able to handle your issues effectively. That's increasingly showing to be not the case.
- rodgerd 11y ago
  IBM is not a company, it's a collection of fiefdoms. Depending on who your contacts are and a whole bunch of Game of Thrones like machinations within the company, they can be wonderful or absolutely abhorrent.
- yellowapple 11y ago
  I'm personally a really big fan of IBM's hardware, particularly their POWER-based systems. IBM hardware simply never dies, and I've witnesses firsthand the "nobody was ever fired for buying IBM" phenomenon. Lenovo's IBM-derived products (i.e. their ThinkPads and related workstations/PCs) are also among my favorites by way of their IBM heritage; ThinkPads in particular are known for being among the best laptops for FOSS operating systems (GNU/Linux, the BSDs, etc.) when it comes to out-of-the-box hardware support, and many of them will even run custom BIOSes like Coreboot. I'm not as much of a fan of their software, though that's probably because I haven't really used a whole lot of it.
  - nstott 11y ago
    the Power stuff is going to be challenging I think, some of our db's don't support that architecture yet
- mark_l_watson 11y ago
  I have used Bluemix and generally like it. I only stopped using it when I was accepted into Microsoft's Bizspark program (Azure has good Linux support and is also easy to use). The days are long over that I would lease raw servers. AWS, Azure, Bluemix, Google Cloud Services, etc. all provide great ancillary services.
- needusername 11y ago
  IBM makes some of the worst, most bloated, buggy and overpriced software I every used: RAD and WAS. They are so bad they are a torture on the developer. There are no technical reasons for choosing them. I get angry just thinking about it.
- nickpsecurity 11y ago
  They're a conglomerate that does three things: sell products; sell services; license software & hardware intellectual property. They made $92 billion doing this last year. They derive this stuff through a combination of several billion a year in R&D, acquisitions of successful companies, and acquisitions of companies with strategically useful I.P.. Unlike most tech companies, they've weathered decades worth of storms. Unfortunately for them, they suffer that big company type of thinking that kills innovation. It's why they've had to acquire more. They're were also late to the cloud market. Their last CEO also practically ran it into the ground to prop up stock price before exiting with almost $300 million personal gain. They're facing hard times and it will be interesting to see how they adapt. Their recent push for innovation and agile is interesting. Hopefully a sign of things to come.
gane5h 11y ago
Congrats to the compose folks! IIRC, they acquired Cloudant a little while ago.
- amyjess 11y ago
  And Softlayer. As a friend of mine who works there says, "IBM wanted to have a great cloud business, so they bought one".
- gt384u 11y ago
  And Blue Box (https://www.blueboxcloud.com/ https://www.blueboxcloud.com/) a month and a half ago. Looks like they're investing in or acquiring cloud platform software companies like gangbusters at the moment.
sagivo 11y ago
> We love being a distributed company. Many of us work from our homes and like to occasionally relocate for weeks or months at a time to work in other parts of the world. I love this working style
bstar77 11y ago
I've been a customer for several years, never once had a problem with reliability or performance. I've used several other services and nothing compared to the simplicity of mongohq/compose. They've also had pretty great support for some obscure issues I've had to deal with. Overall, I hope this move just means more of the same.
tptacek 11y ago
Some of the hardest working people in show business. Congratulations!
- nstott 11y ago
  this makes me happy! thanks
fps 11y ago
This explains a lot. I've been a (large) customer of compose.io for several years, and since the rename to compose support of our MongoDB clusters has really gone downhill. They recently pulled much of the monitoring that was available to us via MMS, they disabled New Relic monitoring for several months with no explanation, and support has been awful for what we're paying - 13 hours for a first response on a CAT-1 issue, due apparently to a corporate outing that took their entire support staff out of the office.
- mrkurt 11y ago
  This is unfortunate. It has nothing to do with the acquisition, but I hate to hear it. New Relic and MMS were previously relatively fragile, hand done addons for some customers. Since the beginning of 2014, we've made an attempt to cut back on snowflakes and build reliable, self service features instead. It doesn't excuse you having a suboptimal experience, I know, but that's what's going on. I'm happy to talk more about things if you want to email me.
- mwj 11y ago
  We have also found their support and general attitude towards us (as a long term customer) getting worse recently, so much so that we are migrating off their platform.
  - mrkurt 11y ago
    Same offer for you! I'm happy to talk over email. I'm sad you're migrating away.
  - putlake 11y ago
    Who are you migrating to? I've been happy with Mongodirector (scalegrid)
    - exhilaration 11y ago
      We use MongoDirector for our production Mongo instances and it's been great. No issues since we switched to them in September 2014.
nickpsecurity 11y ago
Good job. I have an idea that might benefit Compose now. Research projects I've read on in IBM said that IBM allowed the internal use of any of its software tools for free. They have a lot of really good ones that you'd normally pay large sums for. I'm not sure if this policy (a) still exists or (b) applies to companies they acquire. However, if I was Compose, I would ask about it because IBM's tools & tech could greatly aid Compose in developing their stuff faster, more robustly, etc. Plus, I'm sure other startups might want to know whether this benefit exists or not. Might even factor into a decision of whether a company wants to be acquired by IBM vs another company. A 100+ free (or cheap) software products would really really sweeten the deal to me if I was to continue working at the company post-acquisition. Especially those delightful, bug-hunting and productivity-boosting products. :) Anyone on the inside know the answers to any of this?
- LoSboccacc 11y ago
  Just wait bluemix has a track history of building a free tier of every service. Happened with sendgrid i.e. That was the policy at least Bluemix has been weird to me. Got the hardest time to get founding and that annoyed the hell out of me. Couldn't even get space for the blue tilde group running (weird thing, maybe still on w3, maybe not)
- wmf 11y ago
  Using your own software for development is one thing, but building a production SaaS by mooching off other divisions probably isn't sustainable.
  - nickpsecurity 11y ago
    There's risks. There's several approaches that come to mind: straight up free software; software and hardware at cost; one of these for a certain period to allow growth before charging for it. The free software obviously isn't making the company money. However, most internal projects on a budget or startups in this era aren't going to choose IBM's tools. They're too expensive and not that popular. So, by not giving it away, they probably gain nothing in licensing. By giving it away, they gain nothing in licensing but might in the project's goal due to software's advantages. For at cost, it might be sustainable. They continue to make most of their profit outside IBM. Inside IBM, they offer their products anywhere ranging from free to a considerable discount. The offer covers the cost of the product in terms of delivery, support, hardware, and so on. That money fuels development and/or cost-cutting on the offerings in general. Volume of IBM use alone might add several developers to any product. So, in this model, there's no losses past what they currently charge, the products get funds for development, and there are gains throughout organization via their benefits. The final model is basically a trial period. The projects or acquired startups get to use their choice of products with discounts ranging from cheaper to at cost to free depending on budget. As budget or success grow, the discount gets lower. Over time, you have a division that's paying full price for the products or begins transitioning to others. You get the benefits of option 2 with some of what general market might pay. Personally, I think Model 1 wouldn't hurt them except whatever they're charging for internal use. Model 2 seems to be the best with Model 3 having a chance of success. The reason I promote this is that IBM is sitting on a lot more potential than they're using, mostly due to culture. However, I think they could create significant, competitive advantages if all their projects and divisions could inexpensively use every key capability they've produced. Software productivity, security, infrastructure, and administration at the least.
dejv 11y ago
I just hope they are really keeping the pricing on current levels. I am running couple of legacy apps there, but I guess it is better to prepare for migration. Any suggestion on similar services?
- rglover 11y ago
  Not as shiny/pretty, but I've used MongoLab (http://mongolab.com http://mongolab.com) for remote backups and never had any issues.
- jsilvers 11y ago
  Pricing really isn't changing. Hope you stick around!
  - dejv 11y ago
    I am sure I am sticking around as long as possible, but it is always good idea to be ready for all eventualities. Especially when you are running legacy apps that you have no time to update, but still don't want to let them die. Thanks to MongoHQ I was able to move one particular app three times to different providers. It took no time, just upload the code, change DNS and you are good.
- thomcrowe 11y ago
  We're really not changing prices, you're good.
vinhboy 11y ago
Maybe off topic, but why are these hosted search engines so expensive? $45/mo to start. Are there any $5/mo, low volume, machines for hobby developers?
- gk1 11y ago
  Because it's more profitable to sell larger accounts, even if fewer of them, than to sell many cheap accounts. One enterprise user at $100,000/year brings in the same revenue as 1,667 "hobby developers" at $5/mo, but requires significantly less babysitting.
  - brightball 11y ago
    Flip side is that you're less diversified and if you lose an enterprise client you feel it hard. There's more stability with lots of small users if you can make it work.
- mrkurt 11y ago
  Elasticsearch is pretty hefty to run in a basic configuration. Our pricing is a function of the resources we assign to cluster nodes, and a bare minimum Elasticsearch installation needs substantially more power than most non-JVM databases.
- nzadrozny 11y ago
  Relevant to my interests. Founder of ES host Bonsai.io. Echoing mrkurt, ES is indeed hefty to run. Elasticsearch is going to require a hundred+ megs of RAM just to run an empty node. And then you'll want enough heap and CPU that you're not GCing yourself to death while importing data, or while churning through filter and caches. And then enough memory to run those fancy aggregations in Kibana. Times the number of nodes you want for redundancy, and for your master-role quorum. It's possible to do long-tail low-volume ES hosting for hobby development. But to do so you need to dig in and consider ES hosting as a holistic system. We've worked hard at Bonsai to build a hobbyist-friendly hosting system. Our approach is to fork ES itself to introduce native multi-tenancy. That means everyone in the cluster is amortizing the overhead of all that JVM memory (and redundant data nodes, and master-node quorum, a load balancer and authenticating reverse proxy, and analytics... etc...) The problem remains that ES, while easy to spin up, is still a pretty heavy system to manage at scale. So it's generally a lot easier to chuck some extra resources at it, maximize your isolation, and eat the higher entry cost.
kylehotchkiss 11y ago
What makes me most sad is that a really, really great product is being sucked into a company who has no reason to offer services to a person like me. Knowing I could spin up postgres for $18/month was pretty much the sweet spot that gave me the confidence to want to do more personal, outside of work projects. Heroku's $50/month offering is far too much. But I have a feeling that this acquisition, like many others, means the end of the offering of a product to the lone-wolf developer who wants to just build something for fun that won't scale nor make money. I have literally no reason whatsoever to believe IBM cares about the little guy but I'm okay and willing to be proven wrong about that. I want to be proven wrong about that. I don't want an email nor an empty promise. I honestly just want you guys to keep the name "compose" and keep that cheap postgres option open for the long run (years and years not weeks and weeks) so us folks who just want to build something for fun have a quality database with a professional operation to support it at the great pricing it is today.
- detaro 11y ago
  IBM already has some "lone developer" friendly offerings in bluemix, so there is hope
- splatcollision 11y ago
  Just get yourself a free Bluemix account, add the compose DB services when they become available (and they will), and if your usage is on that hobbyist level, you'll only pay what you're paying now. You don't even have to use Bluemix itself for runtime hosting, you can access most Bluemix services from your localhost or other app environments.
  - justincormack 11y ago
    "pretend to sign up to our other services so it looks like IBM is attracting new business and you can continue to use this service for free"
- mrkurt 11y ago
  This is important to us, and was a factor in our decision to go to IBM (which is a little counter intuitive, given IBM's history). Our entire sales model is predicated on people being able to start easily, and grow if they need to. Price, self service-ness, and the content we produce are all important in this model. The thing about lone wolf developers, and side projects, and one off things people hack together over the weekend is that they sometimes "flip" and become big customers. That's when we win, and we monitor our funnel like crazy and continuously see that it's working. Interestingly, some of the bad customer experiences you see in this thread are the direct result of this model. We aren't very good at handling one offs, or doing things by hand, or consulting, because we focus so much on repeatable, self service offerings — which have to work for small applications. IBM is interested in three things from us. The technology is good, the team is great, and the way we sell to customers is something they have told us they firmly believe is the future of selling tools to developers. There's a really, really good chance that you'll see other parts of IBM acting like us (and building things for you) before you see us becoming all enterprisey. We'll keep the name Compose (because it's fabulous!), we'll keep the cheap Postgres option, and we'll do our best to keep that true for years and years. For what it's worth, I greatly appreciate your comment.
rubiquity 11y ago
This is the second "Cloud Database" company IBM has acquired recently. They also acquired Cloudant, a CouchDB based "Cloud Database." Side-thought: Did Compose refer to itself as a Cloud Database before IBM acquired it? That seems like a very IBM marketing oriented term.
cj 11y ago
I'd really love to see a MongoDB provider in between Compose.io and MMS. We were on Compose for 1.5 years, but recently switched off to MMS after we started running in to scaling issues. Compose is a bit limiting if you need to tweak / monitor resources for scaling, but running reliably on MMS means really understanding how your deployment works (which might be the best long-term position to be in, but a bit distracting for a ~5 person team). It'd be great if Compose could offer a managed MongoDB service that ran off instances in my ec2 account (so I'd pay base ec2 rates, plus a $50-150 per-server fee for managing the deployment)
WaltPurvis 11y ago
All of the comments here that talk about price are saying how cheap Compose is. Really? Looks like they charge $12 per gigabyte per month for PostgreSQL, $18 for Mongo, and there's absolutely no free/hobbyist tier. That seems pretty expensive to me, at least compared to the other IBM cloud DBaaS offering, Cloudant, which is priced at $1 per gigabyte per month and is completely free if your bill is less than $50/month. Cloudant charges for transactions (gets and puts), while it looks to me like Compose doesn't, but based on my (admittedly limited) experience with Cloudant that wouldn't do very much to close the yawning gap for most applications. Would someone from Compose care to explain why their pricing isn't really 12X or 18X greater than Cloudant? (And whose pricing is going to change? I'm kind of worried that IBM is going to quadruple Cloudant's prices, and then quadruple them again, to bring them into line with Compose.)
- akhilcacharya 11y ago
  Back when I used Compose/MongoHQ exclusively for my sideprojects, they had a very generous free tier. They only recently switched to stop being generous (which is completely understandable).
- mrkurt 11y ago
  Cloudant's pricing is so different than ours it's hard to compare. Our DBs are all (currently) intended to be used real time, which means they could be handling several thousand requests per second. I would guess that Cloudant's pricing is similar to ours for DBs under load. Theirs just scales down in a way we can't currently. Cloudant's actually going to be really nice for our customers. Since our service and pricing are meant for transactional databases, we've lost customers and potential customers who had mountains of cold data they needed to keep somewhere. As an independent company, that was the right choice, as part of IBM we now have a good answer for those types of data loads. We hope to be able to use our Transporter feature to make it relatively easy to keep data moving around between DBs, Cloudant included.
hangtime79 11y ago
Welcome Compose team to IBM. I sit in analytics tech sales and work with CDS (particularly Dash-DB) pretty regularly and am very excited to have you aboard.
- nstott 11y ago
  thanks, we're all pretty excited
oomkiller 11y ago
Taking a look at Bluemix, I already see most of the Compose offerings listed. RethinkDB is missing though, are you discontinuing that product?

172 points by thomcrowe 11y ago

ukd1 11y ago

Congrats guys!

thomcrowe 11y ago
Pretty exciting!

nfriedly 11y ago

As an IBM'er, I think this is awesome news! Cloudant is awesome for CouchDB, but if you want mongo or redis, we didn't have a lot to offer before today. Looking forward to working with you guys!

compostor42 11y ago

I've very curious what the general "impression" regarding IBM is around HN? They are not discussed much around here.

ahoff 11y ago
IBM is huge, and it's hard to give a general impression around the whole company that makes any sense. I can only really speak about CDS (the division Compose is joining and Cloudant is a part of) but I can say it's a really exciting place to work. CDS is about as fast-moving as you can get in a company this large, and the leadership really understands the shift to SaaS. We have a great team and a lot of autonomy to make decisions outside of the general corporate superstructure. I'm biased but think it speaks highly that Cloudant people (Adam Kocoloski and Derek Schoettle) were tapped to lead this group.
- mbroberg 11y ago
  It's a huge place, but I agree with ahoff, having come over along with the Cloudant acquisition. I feel pretty good about our little corner of IBM in CDS (cloud data services). It's not a bad place to end up if you're at a data management startup.
tdicola 11y ago
IBM is a financial engineering firm, not a technology company anymore: http://dealbook.nytimes.com/2014/10/20/the-truth-hidden-by-ibms-buybacks/?_r=0 http://dealbook.nytimes.com/2014/10/20/the-truth-hidden-by-i...
- ahoff 11y ago
  If you only care about stock price, and even then no. IBM is a large company that does a lot. They put more money and people towards pure tech than most firms, especially in places like IBM Research.
- devy 11y ago
  What people remember them the most, is their R&D. The beat Intel to create the first 7nm chip process recently.
  - derwiki 11y ago
    The Almaden Research Center is still a top-notch research facility.
  - timelined 11y ago
    Between Alamaden (San Jose) and Yorktown Heights (NY) in the US, IBM is one of the last few industrial basic sciences/engineering labs in the US, resembling Bell Labs and Xerox PARC from a few decades ago.
rdtsc 11y ago
I think IBM is turning around. It missed the cloud shift by quite a few years, but they know it, and they are re-inventing themselves, which is quite commendable for a large 100+ year old company. HN often lives in the startup + Facebook + Google + latest tech trend bubble. But there is a whole world of enterprise + government (DoD) tech world out there that exists, but it is not talked much about here, unless it is being made fun of perhaps.
- abutler 11y ago
  IBM is no longer just focused on enterprise and government. There's a big focus now on startups. Take a look at the Bluemix Garage where there's IBM engineers embedded in Galvanize in SF to work with startups but also help enterprises bridge the gap and learn from them; and also look at digital.nyc and tech.london. Both are IBM initiatives. So... IBM today isn't like the IBM that most people imagine and I am sure that the IBM of tomorrow will be completely different. You don't get to be over 100 years old and a tech company unless you can reinvent yourself to stay relevant.
  - rdtsc 11y ago
    That is what I was trying to say, I was just responding to what I think HN crowd's perception is of IBM, which I think is not accurate. I think it is really great to see a company of their size and age change course and adapt.
roymurdock 11y ago
My impression is that they have almost fully divested from hardware and are doing whatever they can to boost their consulting, services, cloud & analytics businesses - hence this latest acquisition, on top of many other cloud purchases. This is smart, given that hardware is a commodity with low and falling margins, whereas it's very hard to price heterogeneous services such as consulting, integration, etc. They've done some shady things to cut costs, but they needed to pivot pretty drastically in order to maintain their position and restart a growth cycle. Ultimately, I think the "big data" market is overhyped, and that IBM is playing into this hype. People are gobbling up data storage and analytics services left and right, but are the valuations equivalent to the value added to a company by storing and sorting a ton of (potentially useless) data?
jdcskillet 11y ago
I just migrated from working directly on the IBM Mainframe in COBOL (gasp) to my true love of .Net and JS (yep, both of them) within the same company. I have been on a number of greenfield projects with IBM and quite frankly, my opinion of them differs based on the area of the company you are dealing with. For instance, I love dealing with the CI guys at Urbancode Deploy, whereas getting similar information from their Rational counterparts (Rational Team Concert) was like pulling teeth. Pretty much every experience I had with the Rational teams was terrible because they would hide information to sell more consulting services or additional products (even though they guaranteed the feature was available on our existing license). I despise the documentation sifting and information gathering required to use IBM products (developer works is a terrible site). However, I have much respect for the engineering that goes into those products and it is a joy when you get to work directly with their engineers to solve a problem. Good luck compose, I hope you get to be autonomous like Urbancode and not stuck in the sand of the Rational teams.
- davidgerard 11y ago
  Rational's pricing model has always been "how much have you got?"
teddyh 11y ago
“That company is large because it is large. […] By not seeking to strive, it conquers without effort.” – The Tao of Programming, chapter 8, Geoffrey James, 1987
fennecfoxen 11y ago
My impression of IBM as a place to work is that it's hopelessly bureaucratic in a way that no simple reform is going to be able to fix. For instance. Once upon a time I had an internship there, working with a particular product dealing with server management. As part of the setup process required to customize servers, users were directed to go to a random university user's home directory to download one particular older version of some open-source boot-utility software (newer versions wouldn't work). This is, of course, a terrifying distribution mechanism. Actually hosting and distributing this file wasn't something that IBM people were willing to do; you'd need to run it through the Lawyers. (This probably feeds a Not Invented Here syndrome.) That said! At least the internship was with the Software division. The way I heard it, Services was 90% of the revenue and Software was 90% of the profit. Things are a lot better in Software. Still a damnable maze of cubicle farms, though.
chris_wot 11y ago
IBM are increasingly pissing off their customers since they retrenched a good portion of their experienced workforce. I have a friend who is a CIO of a large company in Australia and he is increasingly getting annoyed with IBM's SAP support and is hunting around for alternatives. He probably won't move from IBM completely, but they'll be getting less and less business from him. I mean, the other day their entire datacentre's network link was flapping, for hours and hours. It wasn't just his company that was affected, it was affecting even larger companies. IBM's ability to deliver is going down the plughole. They've always been very expensive, but have been able to justify it because they were able to handle your issues effectively. That's increasingly showing to be not the case.
rodgerd 11y ago
IBM is not a company, it's a collection of fiefdoms. Depending on who your contacts are and a whole bunch of Game of Thrones like machinations within the company, they can be wonderful or absolutely abhorrent.
yellowapple 11y ago
I'm personally a really big fan of IBM's hardware, particularly their POWER-based systems. IBM hardware simply never dies, and I've witnesses firsthand the "nobody was ever fired for buying IBM" phenomenon. Lenovo's IBM-derived products (i.e. their ThinkPads and related workstations/PCs) are also among my favorites by way of their IBM heritage; ThinkPads in particular are known for being among the best laptops for FOSS operating systems (GNU/Linux, the BSDs, etc.) when it comes to out-of-the-box hardware support, and many of them will even run custom BIOSes like Coreboot. I'm not as much of a fan of their software, though that's probably because I haven't really used a whole lot of it.
- nstott 11y ago
  the Power stuff is going to be challenging I think, some of our db's don't support that architecture yet
mark_l_watson 11y ago
I have used Bluemix and generally like it. I only stopped using it when I was accepted into Microsoft's Bizspark program (Azure has good Linux support and is also easy to use). The days are long over that I would lease raw servers. AWS, Azure, Bluemix, Google Cloud Services, etc. all provide great ancillary services.
needusername 11y ago
IBM makes some of the worst, most bloated, buggy and overpriced software I every used: RAD and WAS. They are so bad they are a torture on the developer. There are no technical reasons for choosing them. I get angry just thinking about it.
nickpsecurity 11y ago
They're a conglomerate that does three things: sell products; sell services; license software & hardware intellectual property. They made $92 billion doing this last year. They derive this stuff through a combination of several billion a year in R&D, acquisitions of successful companies, and acquisitions of companies with strategically useful I.P.. Unlike most tech companies, they've weathered decades worth of storms. Unfortunately for them, they suffer that big company type of thinking that kills innovation. It's why they've had to acquire more. They're were also late to the cloud market. Their last CEO also practically ran it into the ground to prop up stock price before exiting with almost $300 million personal gain. They're facing hard times and it will be interesting to see how they adapt. Their recent push for innovation and agile is interesting. Hopefully a sign of things to come.

gane5h 11y ago

Congrats to the compose folks! IIRC, they acquired Cloudant a little while ago.

amyjess 11y ago
And Softlayer. As a friend of mine who works there says, "IBM wanted to have a great cloud business, so they bought one".
gt384u 11y ago
And Blue Box (https://www.blueboxcloud.com/ https://www.blueboxcloud.com/) a month and a half ago. Looks like they're investing in or acquiring cloud platform software companies like gangbusters at the moment.

sagivo 11y ago

> We love being a distributed company. Many of us work from our homes and like to occasionally relocate for weeks or months at a time to work in other parts of the world. I love this working style

bstar77 11y ago

I've been a customer for several years, never once had a problem with reliability or performance. I've used several other services and nothing compared to the simplicity of mongohq/compose. They've also had pretty great support for some obscure issues I've had to deal with. Overall, I hope this move just means more of the same.

tptacek 11y ago

Some of the hardest working people in show business. Congratulations!

nstott 11y ago
this makes me happy! thanks

fps 11y ago

This explains a lot. I've been a (large) customer of compose.io for several years, and since the rename to compose support of our MongoDB clusters has really gone downhill. They recently pulled much of the monitoring that was available to us via MMS, they disabled New Relic monitoring for several months with no explanation, and support has been awful for what we're paying - 13 hours for a first response on a CAT-1 issue, due apparently to a corporate outing that took their entire support staff out of the office.

mrkurt 11y ago
This is unfortunate. It has nothing to do with the acquisition, but I hate to hear it. New Relic and MMS were previously relatively fragile, hand done addons for some customers. Since the beginning of 2014, we've made an attempt to cut back on snowflakes and build reliable, self service features instead. It doesn't excuse you having a suboptimal experience, I know, but that's what's going on. I'm happy to talk more about things if you want to email me.
mwj 11y ago
We have also found their support and general attitude towards us (as a long term customer) getting worse recently, so much so that we are migrating off their platform.
- mrkurt 11y ago
  Same offer for you! I'm happy to talk over email. I'm sad you're migrating away.
- putlake 11y ago
  Who are you migrating to? I've been happy with Mongodirector (scalegrid)
  - exhilaration 11y ago
    We use MongoDirector for our production Mongo instances and it's been great. No issues since we switched to them in September 2014.

nickpsecurity 11y ago

Good job. I have an idea that might benefit Compose now. Research projects I've read on in IBM said that IBM allowed the internal use of any of its software tools for free. They have a lot of really good ones that you'd normally pay large sums for. I'm not sure if this policy (a) still exists or (b) applies to companies they acquire. However, if I was Compose, I would ask about it because IBM's tools & tech could greatly aid Compose in developing their stuff faster, more robustly, etc. Plus, I'm sure other startups might want to know whether this benefit exists or not. Might even factor into a decision of whether a company wants to be acquired by IBM vs another company. A 100+ free (or cheap) software products would really really sweeten the deal to me if I was to continue working at the company post-acquisition. Especially those delightful, bug-hunting and productivity-boosting products. :) Anyone on the inside know the answers to any of this?

LoSboccacc 11y ago
Just wait bluemix has a track history of building a free tier of every service. Happened with sendgrid i.e. That was the policy at least Bluemix has been weird to me. Got the hardest time to get founding and that annoyed the hell out of me. Couldn't even get space for the blue tilde group running (weird thing, maybe still on w3, maybe not)
wmf 11y ago
Using your own software for development is one thing, but building a production SaaS by mooching off other divisions probably isn't sustainable.
- nickpsecurity 11y ago
  There's risks. There's several approaches that come to mind: straight up free software; software and hardware at cost; one of these for a certain period to allow growth before charging for it. The free software obviously isn't making the company money. However, most internal projects on a budget or startups in this era aren't going to choose IBM's tools. They're too expensive and not that popular. So, by not giving it away, they probably gain nothing in licensing. By giving it away, they gain nothing in licensing but might in the project's goal due to software's advantages. For at cost, it might be sustainable. They continue to make most of their profit outside IBM. Inside IBM, they offer their products anywhere ranging from free to a considerable discount. The offer covers the cost of the product in terms of delivery, support, hardware, and so on. That money fuels development and/or cost-cutting on the offerings in general. Volume of IBM use alone might add several developers to any product. So, in this model, there's no losses past what they currently charge, the products get funds for development, and there are gains throughout organization via their benefits. The final model is basically a trial period. The projects or acquired startups get to use their choice of products with discounts ranging from cheaper to at cost to free depending on budget. As budget or success grow, the discount gets lower. Over time, you have a division that's paying full price for the products or begins transitioning to others. You get the benefits of option 2 with some of what general market might pay. Personally, I think Model 1 wouldn't hurt them except whatever they're charging for internal use. Model 2 seems to be the best with Model 3 having a chance of success. The reason I promote this is that IBM is sitting on a lot more potential than they're using, mostly due to culture. However, I think they could create significant, competitive advantages if all their projects and divisions could inexpensively use every key capability they've produced. Software productivity, security, infrastructure, and administration at the least.

dejv 11y ago

I just hope they are really keeping the pricing on current levels. I am running couple of legacy apps there, but I guess it is better to prepare for migration. Any suggestion on similar services?

rglover 11y ago
Not as shiny/pretty, but I've used MongoLab (http://mongolab.com http://mongolab.com) for remote backups and never had any issues.
jsilvers 11y ago
Pricing really isn't changing. Hope you stick around!
- dejv 11y ago
  I am sure I am sticking around as long as possible, but it is always good idea to be ready for all eventualities. Especially when you are running legacy apps that you have no time to update, but still don't want to let them die. Thanks to MongoHQ I was able to move one particular app three times to different providers. It took no time, just upload the code, change DNS and you are good.
thomcrowe 11y ago
We're really not changing prices, you're good.

vinhboy 11y ago

Maybe off topic, but why are these hosted search engines so expensive? $45/mo to start. Are there any $5/mo, low volume, machines for hobby developers?

gk1 11y ago
Because it's more profitable to sell larger accounts, even if fewer of them, than to sell many cheap accounts. One enterprise user at $100,000/year brings in the same revenue as 1,667 "hobby developers" at $5/mo, but requires significantly less babysitting.
- brightball 11y ago
  Flip side is that you're less diversified and if you lose an enterprise client you feel it hard. There's more stability with lots of small users if you can make it work.
mrkurt 11y ago
Elasticsearch is pretty hefty to run in a basic configuration. Our pricing is a function of the resources we assign to cluster nodes, and a bare minimum Elasticsearch installation needs substantially more power than most non-JVM databases.
nzadrozny 11y ago
Relevant to my interests. Founder of ES host Bonsai.io. Echoing mrkurt, ES is indeed hefty to run. Elasticsearch is going to require a hundred+ megs of RAM just to run an empty node. And then you'll want enough heap and CPU that you're not GCing yourself to death while importing data, or while churning through filter and caches. And then enough memory to run those fancy aggregations in Kibana. Times the number of nodes you want for redundancy, and for your master-role quorum. It's possible to do long-tail low-volume ES hosting for hobby development. But to do so you need to dig in and consider ES hosting as a holistic system. We've worked hard at Bonsai to build a hobbyist-friendly hosting system. Our approach is to fork ES itself to introduce native multi-tenancy. That means everyone in the cluster is amortizing the overhead of all that JVM memory (and redundant data nodes, and master-node quorum, a load balancer and authenticating reverse proxy, and analytics... etc...) The problem remains that ES, while easy to spin up, is still a pretty heavy system to manage at scale. So it's generally a lot easier to chuck some extra resources at it, maximize your isolation, and eat the higher entry cost.

kylehotchkiss 11y ago

What makes me most sad is that a really, really great product is being sucked into a company who has no reason to offer services to a person like me. Knowing I could spin up postgres for $18/month was pretty much the sweet spot that gave me the confidence to want to do more personal, outside of work projects. Heroku's $50/month offering is far too much. But I have a feeling that this acquisition, like many others, means the end of the offering of a product to the lone-wolf developer who wants to just build something for fun that won't scale nor make money. I have literally no reason whatsoever to believe IBM cares about the little guy but I'm okay and willing to be proven wrong about that. I want to be proven wrong about that. I don't want an email nor an empty promise. I honestly just want you guys to keep the name "compose" and keep that cheap postgres option open for the long run (years and years not weeks and weeks) so us folks who just want to build something for fun have a quality database with a professional operation to support it at the great pricing it is today.

detaro 11y ago
IBM already has some "lone developer" friendly offerings in bluemix, so there is hope
splatcollision 11y ago
Just get yourself a free Bluemix account, add the compose DB services when they become available (and they will), and if your usage is on that hobbyist level, you'll only pay what you're paying now. You don't even have to use Bluemix itself for runtime hosting, you can access most Bluemix services from your localhost or other app environments.
- justincormack 11y ago
  "pretend to sign up to our other services so it looks like IBM is attracting new business and you can continue to use this service for free"
mrkurt 11y ago
This is important to us, and was a factor in our decision to go to IBM (which is a little counter intuitive, given IBM's history). Our entire sales model is predicated on people being able to start easily, and grow if they need to. Price, self service-ness, and the content we produce are all important in this model. The thing about lone wolf developers, and side projects, and one off things people hack together over the weekend is that they sometimes "flip" and become big customers. That's when we win, and we monitor our funnel like crazy and continuously see that it's working. Interestingly, some of the bad customer experiences you see in this thread are the direct result of this model. We aren't very good at handling one offs, or doing things by hand, or consulting, because we focus so much on repeatable, self service offerings — which have to work for small applications. IBM is interested in three things from us. The technology is good, the team is great, and the way we sell to customers is something they have told us they firmly believe is the future of selling tools to developers. There's a really, really good chance that you'll see other parts of IBM acting like us (and building things for you) before you see us becoming all enterprisey. We'll keep the name Compose (because it's fabulous!), we'll keep the cheap Postgres option, and we'll do our best to keep that true for years and years. For what it's worth, I greatly appreciate your comment.

rubiquity 11y ago

This is the second "Cloud Database" company IBM has acquired recently. They also acquired Cloudant, a CouchDB based "Cloud Database." Side-thought: Did Compose refer to itself as a Cloud Database before IBM acquired it? That seems like a very IBM marketing oriented term.

cj 11y ago

I'd really love to see a MongoDB provider in between Compose.io and MMS. We were on Compose for 1.5 years, but recently switched off to MMS after we started running in to scaling issues. Compose is a bit limiting if you need to tweak / monitor resources for scaling, but running reliably on MMS means really understanding how your deployment works (which might be the best long-term position to be in, but a bit distracting for a ~5 person team). It'd be great if Compose could offer a managed MongoDB service that ran off instances in my ec2 account (so I'd pay base ec2 rates, plus a $50-150 per-server fee for managing the deployment)

WaltPurvis 11y ago

All of the comments here that talk about price are saying how cheap Compose is. Really? Looks like they charge $12 per gigabyte per month for PostgreSQL, $18 for Mongo, and there's absolutely no free/hobbyist tier. That seems pretty expensive to me, at least compared to the other IBM cloud DBaaS offering, Cloudant, which is priced at $1 per gigabyte per month and is completely free if your bill is less than $50/month. Cloudant charges for transactions (gets and puts), while it looks to me like Compose doesn't, but based on my (admittedly limited) experience with Cloudant that wouldn't do very much to close the yawning gap for most applications. Would someone from Compose care to explain why their pricing isn't really 12X or 18X greater than Cloudant? (And whose pricing is going to change? I'm kind of worried that IBM is going to quadruple Cloudant's prices, and then quadruple them again, to bring them into line with Compose.)

akhilcacharya 11y ago
Back when I used Compose/MongoHQ exclusively for my sideprojects, they had a very generous free tier. They only recently switched to stop being generous (which is completely understandable).
mrkurt 11y ago
Cloudant's pricing is so different than ours it's hard to compare. Our DBs are all (currently) intended to be used real time, which means they could be handling several thousand requests per second. I would guess that Cloudant's pricing is similar to ours for DBs under load. Theirs just scales down in a way we can't currently. Cloudant's actually going to be really nice for our customers. Since our service and pricing are meant for transactional databases, we've lost customers and potential customers who had mountains of cold data they needed to keep somewhere. As an independent company, that was the right choice, as part of IBM we now have a good answer for those types of data loads. We hope to be able to use our Transporter feature to make it relatively easy to keep data moving around between DBs, Cloudant included.

hangtime79 11y ago

Welcome Compose team to IBM. I sit in analytics tech sales and work with CDS (particularly Dash-DB) pretty regularly and am very excited to have you aboard.

nstott 11y ago
thanks, we're all pretty excited

oomkiller 11y ago

Taking a look at Bluemix, I already see most of the Compose offerings listed. RethinkDB is missing though, are you discontinuing that product?