CMIPS is a project created by Carles Mateo (blog) LinkedIn.
Being Entrepreneur, Developing Software since 5 y.o., after working with the Cloud* for years, and missing information comparing performance and price between Cloud providers, he decided to create his own tool to measure the performance of the different Cloud Instances and share that info with the Community and the Start ups.
The source code is released as Open Source, so everyone can do their own tests.
Right now you can download the binaries for CMIPS v.1.0.5 (Linux 64 bits) for benchmarking CPU + RAM Access from github:
git clone https://github.com/cmips/cmips_bin
And the source code + binary for CMIPS v. 1.0.5 (CPU + RAM Acces benchmarks)
git clone https://github.com/cmips/cmips
Disclaimer: All the opinions on performance, Services provided by Cloud providers, and everything else are my views only. About performance all the information and results are my tests and are provided without warranty. (However you are free to do your own tests and verify the data)
If you own a Cloud provider that has not been evaluated and you want to CMIPS it, do not hesitate to contact me. If you want a consultancy to find bugs and suggest features send me a LinkedIn’s InMail.
If you are a Startup asking for advice or you’re interested in certain Cloud provider being tested, just drop me a line.
* Experience on the Cloud: Carles had an Start up in 2004 and in addition to having everything virtualized in his Servers, he was selling Virtual Machines to other Start ups (before the terms VPS, Instance, or tenant was created) with Vmware.
Carles was tech lead at Easy Cloud Managed where he created auto-scaling (for Clouds not supporting auto-scaling as well), abstraction mechanisms for deploying to multiple OS (Linux and windows) and different Cloud providers, and other wonderful stuff. So he knows about how many of them internally work: Amazon, OpenStack, Linode, rackspace, eucalyptus, digital ocean… their API’s, tools, libraries like libcloud, orchestration mechanisms and Software and automation like cloud-init, user-data…
He also helped to benchmark, fix some problems and improve google cloud.
He worked for several ISPs as Responsible/SysAdmin/DevOps, and has used many other ISP and Cloud providers while being CTO, Developer and while helping companies with their problems, so knows a wide range of providers and solutions from them, like OVH. So he saw very good things and very very bad.
Carles loves to visit Data Centers and he doesn’t miss a change to visit new. He has worked in many of them and also has visited some of the Fortune 500 (sorry it is confidential, but they are very cool). Has seen Data Centers from multinationals he has worked for (Volkswagen, winterhur insurance..), commercial ones like equinix, colt telecom, phone companies, etc… across several countries in Europe and US. He also setup and created the Data infrastructure for BBV bank (now BBVA bank) ticketing division.
Carles specially hates shared hosting and companies that offer shared hosting and kill processes and MySql instances depending on the CPU use quota (some of them put the email services also under the same CPU quota, so if 10 people at the office download email at the same time, the web stops working!!. Yes, even if it seems unbelievable, bad services like that exist in the not-Enterprise world). He has helped many Developers with crap hostings having problems with Drupal and WordPress installations.
He also used RightScale while leading at some Start ups, mainly for auto-scaling Web Servers with Amazon.
Carles collaborates regularly with several Cloud providers like CloudSigma or AckStorm Cloud. And was invited by Amazon to Madrid, where he was finalist in 2016 for being the SDM of reference in the South of Europe, for the new Development Center, and was invited to Dublin, where he was finalist in the process for being SDM of the CloudWatch project in year 2013 (he was told that the project was so easy for him that he’ll get bored). He really enjoyed meeting so cool Engineers there, so he shared many ideas to improve Amazon’s services, many of them were implemented.
Carles also was invited by Facebook to Menlo Park, Silicon Valley, and finalist for a SDM (Software Devlopment Manager) in the SRE process. This was about physical Servers, not cloud, but had a lot of fun discussing about extreme scaling. After the interview one top Engineer managing tens of thousands of Servers congratulated him and told him that he has learned several new tricks that they will use to upgrade entire racks more quickly. 😉
He also has contributed articles to blogs from CloudSigma and LunaCloud.
He is advisor of several companies and CTO of several growing Start ups.
He also loves Software Development, and Scaling but this comes in another website. 🙂