As of a few minutes ago, Galileo RC3 EPP packages are available from the Eclipse website. You can download them via a direct link to the Galileo RC3 packages website or by opening the main Eclipse download page.  Note, if you go that way, you need to switch to the “Develompent Builds” tab.

Why am I writing this? The main reason is that we – myself and all the other package maintainers – like to get feedback about the packages. This year, we provide even more packages! There will be a new PHP package (the name is self explanatory) and the Pulsar package which contains the bits from the Pulsar Working Group.

Back to the main topic: Feedback. If you want to send us bug reports about the packages, you can do so by pressing the “more…” link on one of the download pages:

This will bring you to a page where you can find in-depth information about every package. On the right-hand side, you will find a section with download links, MD5 and SHA1 checksums, but also relevant Bugzilla queries and a direct link that allows you to send bug reports about a particular package.

But before opening a new bug, please read through the list of open bugs to avoid duplicates and consider filing the bug against a particular project if its related to a project and not to the package itself.

…and the logo that appears on every eclipse.org web page reminds me again: Galileo arrives in just a few weeks!

Tags: eclipse, epp, galileo

We are all busy preparing our talks and demonstrations for EclipseCon, aren’t we? This year I am trying to use in all of my talks a common set of examples.

The plan is to showcase technology from several projects. I will start with the EPP Download Wizard (at the moment ‘Friends-of-Eclipse’ only) from my EPP project that has been build as a RAP application and uses Equinox, p2, and the p2 installer. With the EPP Download Wizard, a user is able to pick and choose from a set of available installable units. On the last wizard page, he or she can download a small customized p2 installer that will download the selected pieces from the p2 repositories. There will be various talks about the things that we do in EPP, e.g. the talk of my colleagues about the internals of the EPP Wizard, or my other talk about the EPP Usage Data Collector.

But what if a developer who is developing this wizard or another application wants to deploy it on a server? The first problem: You need a server! The answer to this question is easy: Use g-Eclipse, our Grid and Cloud Computing tooling, and start your own server on the cloud. This can be done within minutes and we are going to show this in our Equinox above the Cloud talk.

Now that you have your own server up and running, all you need to do is deploying your application on it. But how is this done… well, attend our talk and you will find out how easy it is with p2!

Tags: aws, cloud computing, ec2, ecf, eclipsecon, epp, g-Eclipse, OSGi, p2, rap

I have to admit: 23 is the number that I like most. Some computer scientists and mathematicians prefer 42, but this is not a prime number.

Why am I starting this blog entry with this odd number? Because my birthday is on a 23rd? While this is true, it is not particular interesting. But if you are going to the EPP project page where you see the EPP package builds for all Ganymede Releases you will find out that every Ganymede build that has been released to the public in the past was in fact build on a 23rd.

Look at the screenshot: Ganymede Release build on June, 23rd – Ganymede ServiceRelease 1 on September, 23rd, and now the Ganymede Service Release 2 build on February, 23rd.

And now I am happy that I can write: The new Ganymede SR2 packages are now available for download. There are still some problems on the different download web pages but I am sure those problems will be fixed by the webmasters soon. I think this is good news at the end of a long week.

BTW: How many Eclipse projects did participate in Ganymede? Yes, exactly 23!

Tags: eclipse, epp, ganymede

The Eclipse Packaging Project is working on a wizard that allows you to build your own customizable download, combining plug-ins from predefined packages and Ganymede. The wizard uses Eclipse runtime technology – namely RAP for the web frontend and Equinox P2 for the installer. Before exposing the wizard to a wider audience, we had one burning question: Does it scale up to the volume of downloads being served by Eclipse.org? Read on to find out.

Load Test – Setup

We conducted a load test on the wizard and would like to share the results with you.

The tests were executed on a modern desktop computer with a Core2Duo 2.8 GHz CPU running Ubuntu Hardy. The JVM used a maximum of 512 MB of heap space and ran the concurrent marksweep garbage collector (-UseConcMarkSweepGC). The tests were run using JMeter 3.2 with 200 concurrent threads. Tomcat 6 was used for the server.

Load Test – Results

The results are promising: with a throughput of 100 requests per second we achieved an average response time of 14 miliseconds. Configuring a download takes up to 80 requests, which means we can serve approximately 100,000 downloads per day with a similar configuration.

Here’s a JMeter screenshot with more details on the results:

The fast average response time is paired with a low standard deviation of 60-80 milliseconds. The maximum response times were caused by full garbage collections, but only a few users should experience these response times (otherwise we would have higher average response times).

The allocation of heap is visualized in the diagram below, after multiple hours of load testing. A heap space of 250 MB was constantly needed for the approximately 1,000 concurrent user sessions that were simulated (250KB per session). Another 200 MB was built up and collected every one to two minutes.

Cloud Tests

To demonstrate performance we also ran tests against the same application running on the smallest available type of Amazon EC2 instance.
We ran the same load test as before – but using only 50 concurrent threads. After a period of 16 hours 16,000 sessions had been served, which means a throughput of about 24,000 sessions a day. This is almost equivalent to all the downloads served from Eclipse.org on a busy day. Average response time was approximately 100ms, with network latency accounting for most of the response time. CPU utilization on the server was around 15%.

Conclusion

We’re very positive about the results so far. Even though there is a memory overhead associated with RAP applications, in practice the performance is very solid. Random sampling of the application via a browser in parallel with the load tests showed that responsiveness was very good.

But don’t just take our word for it – try it out for yourself. The application is hosted on an EC2 instance reachable at
http://eppwizard.dyndns.org/eppwizard/rap.
The load test is still running with 50 concurrent sessions against that same instance.

We plan to beta test the wizard with the “Friends of Eclipse” after seeking initial feedback from the community on both usability and performance.
Take it for a spin and tell us what you think.

Tags: epp, load, performance, rap