Anyone designing a high performance transaction processing system has already needed at least a caching solution for increasing response time on frequently used data. But sometimes these data might become bigger than usual, those are times when you need a sofisticate

WebSphere eXtreme Scale Book

solution, something that can automatically fetch data from a slower storage, offload when not necessary, split its contents between various nodes, …

If you require something like this then you’ll notice that you need a DataGrid solution.

By now you’ll be asking yourself why I am on that subject? Recently I’ve been kindly invited to review a book about IBM WebSphere eXtreme Scale from Packt Publishing.

Apart from this, I work on a project where we were already evaluating WebSphere eXtreme Scale as a replacement for WebSphere DynaCache.

If you are in a hurry and want to have a sneak peak on the book contents, take a look at the sample chapter (Chapter 7: “The DataGrid API”) available for free on Packt Publishing website.

For sysadmins this might be a rather easy task, but as a HMC end-user I had to search a little bit.
If you never used an IBM server you might be wondering what is an HMC. At first it seems like a notebook in a blade format as you can see on the picture below:

HMC (source: http://www.fz-juelich.de/jsc/index.php?index=1979)

Setup SSH tunnel and Download JNLP

Open the SSH connection between you and any host on the same network as your HMC (this can even be the HMC itself) and tunnel the following ports from your machine to HMC’s IP.

• 30000 up to 30009
• 9090
• 9960
• 8443
• 443
• 80

Sincerely, I did not check if, strictly speaking, all the ports are necessary but feel free to test them and leave  a comment. These ports were the result of a few searches over the internet.

After setting up the tunnel, point your preferred browser to http://localhost and download the JNLP file at the following link:

HMC html page

I had to edit the codebase property of the JNLP file, but I guess it wont be necessary as I still had to edit the hosts file and add an alias for the machine name (eg.: hmc.my-server.mylocalnetwork).  You’ll end up having codebase as follows:

codebase="http://127.0.0.1/wsmship/codebase"

Edit hosts file

The next step is to edit your hosts file so it will have an extra alias pointing to 127.0.0.1. Copy the host name from the original value of the codebase property.
Then launch the JNLP file with Java Web Start. If you have the mime type correctly associated at your operating system, double click the file and wait for its launch. Sometimes it may complain at the first connection attempt, retry it and it’ll connect successfully. I guess I might still be missing some ports from the 30000~30009 range, since my first attempt was using ports from 30001~30009 and it always complained at the first attempt.

IBM has a rather new platform for messaging. It is called WebSphere MQ Low Latency Messaging (aka WebSphere MQ LLM). It was released on the fourth quarter of 2007 and sincerely it hasn’t much similarities with the well known WebSphere MQ (former MQ Series).

First of all, after installing it you will probably notice that there isn’t a service for starting up. Soon you notice that it is rather a library that provides messaging services. Actually, it is a native library (currently [v2.1] only available for Linux x86, Windows x86 and Solaris Sparc) that comes already with JNI bindings.

To make things even harder, there isn’t much documentation available (and it seems like IBM is still organizing its docs online).

But, IBM claims that it has shown impressive numbers on its benchmarks:

WebSphere MQ Low Latency Messaging has demonstrated very high throughput, one-to-many multicast messaging, which can deliver approximately one million 120-byte messages per second on Ethernet, close to three million 120-byte messages per second on InfiniBand, and more than 8 million smaller messages per second, all on common x86 servers. Testing has also measured very low latency of 30 microseconds for 120 byte messages delivered at 10,000 messages per second on InfiniBand or 61 microseconds on Ethernet (1)

Source: IBM press release

In its Java version, the main classes that you’ll interact when using WMQ LLM are:

• RUMInstance (available through RUMFactory)
• RMMInstance

For instance, RUM stands for Reliable Unicast Messaging and RMM stands for Reliable Multicast Messaging. As you can guess, RUM is used for Queue styled messaging and RMM for Topic styled messaging.

One of its drawbacks is that it still lacks a JMS based Resource Adapter but nothing denies you from rolling your own if you can’t wait for an official adapter from IBM.

I am planning to do some benchmarking against a regular MQ integrated application. As soon as I have the results I’ll post them over here.