Translator’s note: This is the 22nd, and the final article of the series “Cloud Foundry 100-day Challenge selection”. “#100+1” in the title means that it is one extra post (published on November 16, 2015) of the original 100-article series in Japanese. Thank you for reading!
Original Author: Noburou TANIGUCHI (GitHub) (BitBucket)
In Closing
The “Cloud Foundry 100-Day Challenge”, kicked off on June 4th, 2015, has finally come to a close. This was a project to attempt deploying open source applications onto Cloud Foundry, an open source PaaS, over the course of 100 days. Although we took a 3-week break, we were able to complete the project without missing any deadlines, thanks to the various authors, contributors, readers, supporters, and everyone else that helped in making this a success.
In closing, we would like to highlight some statistics regarding this project.
Successes/Failures
First of all, we were able to successfully deploy 97 applications, while we failed to deploy 3. For those of you interested in the failures, here are the links (Translator’s note: All in Japanese):
The cause for “failures” varies in each case. But particulary about Milkode, because it runs on Heroku (Translator’s note: In Japanese), we believe that it were simply matters of not having enough time/skills; a little bit more of both should have allowed us to succeed in deploying it.
Of course, I should mention that all the “successes” do not necessarily represent that we were able to confirm the flawless operation of all functions.
Buildpacks
Next, we tallied the usage of buildpacks for the deployment of the applications, by type.
| Buildpack | Usage |
|---|---|
| Go | 4 |
| Java | 8 |
| Node.js | 22 |
| Null | 1 |
| PHP | 22 |
| Python | 3 |
| Ruby | 21 |
| Staticfile | 16 |
| https://github.com/Altoros/jenkins-buildpack | 1 |
| https://github.com/ddollar/heroku-buildpack-apt | 4 |
| https://github.com/ddollar/heroku-buildpack-multi | 6 |
| https://github.com/ihuston/ipython-notebook-buildpack | 1 |
| https://github.com/nota-ja/heroku-buildpack-hugo | 1 |
| https://github.com/nota-ja/heroku-buildpack-ruby-jekyll | 1 |
| https://github.com/AdmitHub/meteor-buildpack-horse | 2 |
| Total | 113 |
The Total adds up to more than 100, as some applications necessitated the use of multiple buildpacks. Node.js, PHP, and Ruby were the top three, each with over 20, and Staticfile followed. As for Java, Python, and Go, we would have liked to work with them a little more.
Services
The usage of services, by type, looked like this:
| Service | Usage |
|---|---|
| MongoDB | 8 |
| MySQL | 27 |
| MySQL (ClearDB) | 1 |
| PostgreSQL | 12 |
| Redis | 3 |
| SendGrid | 1 |
| Total | 55 |
52 applications, approximately half of the total, used services, and three of which used multiple services. As can be seen, MySQL was by far the most popular, followed by PostgreSQL. MongoDB was surprisingly popular, and by contrast, Redis did not come up as much as expected.
Taking a Look Back at the Initial Objectives
Lastly, we would like to compare and evaluate our achievements against the three initial objectives that we mentioned in #000 Preamble . It is no “numerical” analysis, but more of a reflection of sorts.
- To familiarize users with the features of Cloud Foundry
Access counts for the blog increased significantly from before the launch of the 100-Day Challenge, but the absolute figure (very confidential) could have been higher.
We feel the need to do better. - To understand what points to pay attention to when running applications on Cloud Foundry
We believe that this was achieved at quite a high level among the authors of the articles.
Personally, I feel that I have become better at making predictions on to make various applications work. - To understand the needs and demands of Cloud Foundry users
We unfortunately were not able to receive too many responses to our articles, but in deploying these applications, we were able to form our own opinions on the areas where we would like to see improvements going forward.
In summary, although we do find room for improvement in the fulfilment of our objectives, this does complete, for now, our 100-Day struggle. Again, we thank our authors, contributors, supporters, readers, and everyone else involved; your contributions are highly appreciated.
