Curiosities abound at MXDJ

Steve Erat - Industry Blog ... Really?

Looking at my blog referrers, I found this one popping up quite frequently today:

http://mxdj.sys-con.com/read/rss/100.htm


When I clicked on this link I was taken to a blog aggregator that displayed a list of images in the left menu, and to my surprise, there was a rather low-res graphic of me, with a banner "Steve Erat - Industry Blog". Really? Industry blog? Hadn't thought of it that way before. To me its just my personal blog. In fact, it says so on my TalkingTree banner image.

[More]

More photos of Spain




El Solano and Valle de Benasque of Huesca, in Aragon, Spain
Arasan Liri
Eresue &
Ramastue
Benasque &
Ancilles &
Cerler
Eresue Valle de Benasque
Arasan Liri


Barcelona's Laberint, and Holiday Festivities
Laberynth / Laberint Celebrations / Trobades
Maze /
Laberint
Celebrations /
Trobades

Solution: java.lang.NoClassDefFoundError: com/zerog/lax/LAX

This week I've been setting up my new PowerMac G5, and in the course of preparing to install ColdFusion on JRun I hit a wall in the form of a ZeroG installer error when installing Updater 5 for JRun.

[More]

The Best of Concord Police Logs

I love reading the crazy things that happen around the sleepy town of Concord, as reported in the weeekly Police Logs from the Concord Journal. I've lived in a lot of places, some where most people wouldn't bat an eye if a carjacking were happening, but here in Concord the residents are easily spooked... Enjoy!

I'm proud to claim my own place in the town's log. Here's the what ended up in the newspaper when I was spotted taking pictures in the early morning hours:

    Sunday, Oct. 16 2005
  • At 1:23 a.m., officer on patrol reported a male subject on foot in the Milldam. Officer spoke with the individual, who was taking photos for a college course.


This blog entry will be updated with a selections taken from those logs.



    Week of June 12, 2006
  • At 11:45 a.m., a Simon Willard Road caller reported someone had dumped a bag of brush in her yard. (Why would someone call the police for this? Just pick up the bag and put it in the trash!)
  • At 11:38 a.m., staff from a Sudbury Road country club asked for police assistance with a squirrel that was stuck on a roof. The fire department was dispatched.
  • At 6:13 a.m., emergency crews responded to reports of a smoldering bonfire at White Pond.
  • At 11:48 a.m., a cellular caller reported a motor vehicle versus deer accident on Shadyside Avenue. The deer left the scene.


    Week of June 5, 2006
  • At 7:58 p.m., police received a call about a trespassing bather and his children at White Pond. When confronted the bather apparently got agitated and used profanity.
  • At 8:35 p.m., a Lexington Road caller reported a loose horse and carriage heading down the street. The horse was wrangled by its owner.
  • At 10:31 a.m., a Peter Bulkeley Road caller reported a three-legged pit bull and a Welch corgi running through her yard.
  • At 1:04 p.m., a Manuel Drive caller reported a former friend had posted her picture on a Web site. The pictures were not obscene, the caller said, nor were they appreciated. Police told the caller to contact police in Shirley, where her former friend resides.
  • At 5:19 p.m., a guest at an Elm Street hotel reported she was concerned about the children in the room above her because they had stopped making noise. In an apparently unrelated incident, she also reported heavy set male walking in the hallway. Then she asked for a phone number for the CIA because the FBI had been rude to her.
  • At 7:42 p.m., a resident walked into the police station with her son to report someone had sent the son unwanted instant messages while he as on the computer.


    Week ending July 25, 2005
  • At 8:45 p.m., a College Road caller reported finding dog feces in her mailbox.
  • At 2:50 p.m., a caller reported a male drinking a beer in a vehicle at the West Concord Plaza.
  • At 2:59 p.m., a caller reported a couple engaged in a "public display of affection" on a bench near The Concord Depot.
  • At 5:58 p.m., a Shore Drive caller reported youths swimming in White Pond.
  • At 1:21 a.m., a Shore Drive caller reported people being loud in the area of White Pond. Police found no one at the beach and placed a pair of sneakers, two pairs of sandals, a cell phone, three shirts and a belt into property.
  • At 7:28 p.m., a caller reported two vehicles weaving in and out of traffic shooting water guns at each other. State Police responded

[More]

J2EE Clustering with Tangosol's Coherence

Last night I attended the local Boston Java Meetup Group to get a sense of what my technology cousins are up to since most of my conversations are always in Macromedia-centric arenas.

The most interesting part of the evening for me was meeting Cameron Purdy. Cameron was obviously the Java guru at this table of Java gurus. It turns out that Cameron is president of Tangosol, a company in Somerville which he founded. Tangasol's flagship product Coherence is a type of J2EE clustering software which appropriate for storing extremely large quantities of data in memory and replicating it across large numbers of J2EE cluster members while handling thousands of data updates per second.

Coherence is described in various contexts as:

Tangosol Coherence enables in-memory data management for clustered J2EE applications and application servers. Coherence makes sharing and managing data in a cluster as simple as on a single server. It accomplishes this by coordinating updates to the data using cluster-wide concurrency control, replicating and distributing data modifications across the cluster using the highest performing clustered protocol available, and delivering notifications of data modifications to any servers that request them.


and

... provides scalable performance and is not compromised by single points of failure ... [and] manages many gigabytes of data and thousands of updates per second ... self partioning architecture without single points of failure or measureable GC pauses ...


That last quote is about as much as I could jot down while watching this presentation on Coherence while trying to keep up with Cameron's break-neck reading velocity. As an aside, this presentation was hauntingly similar to a Breeze presentation, but actually used a similar product from Articulate.

This large scale clustering software appears to be significantly more robust than the JINI-based clustering found in Macromedia JRun. Replication of session data across small JRun clusters is generally thought to have a noticible performance impact and its typically recommended to minimize the data that is replicated in this manner to unique keys that act as pointers to persistant data on the backend databases. A couple reviews of JRun clustering are available for comparison:

Roses are red, blogs are blue

As a Valentine's Day gift, I spent this weekend building a new blog for my wife based on Ray Camden's BlogCFC, merged with the Aura skin.

The new blog is called Casa ArbaÃ’il, the name of her family's second home in a small village up in the Pyrenees of Spain, where my wife spent many summers growing up and where we continue to travel each Christmas. As a native of Barcelona, a Catalan speaker, a current resident of Massachusetts, and a chemist by profession you'll find future blog entries on these categories. As soon as her we get her domain name www.casa-arbanil.com mapped to our ISP then the blog will located there, until then it remains on TalkingTree.com.

[More]

Keep Library Loaded - Check this box to retain the library in RAM

I just worked a case where a ColdFusion MX 6.1 server was restarting over and over again every 1-5 minutes with virtually no helpful information in any of the ColdFusion, JRun, Webserver, or Windows Event logs, and there wasn't a HotSpot crash log anywhere.

The app was being tested in staging using the Microsoft Web Stress Tool, which will just hammer one or more designated pages. Running a brief stress test on an isolated test page having only "hello world" did not cause the CFMX server to restart, running the test on the application's main index.cfm easily crashed the server.

The app was loaded onto another CFMX server and the CFAdmin was adjusted for the CFX tags and the datasources. The same stress test did not crash the CFMX server here. A detailed comparison of the two servers and the two CF Admins was made, and it was noticed that the crashing server did not have the checkbox enabled to keep the C++ CFX tag loaded in memory, whereas the stable server was checked.

On the detail view for "Add/Edit C++ CFX Tag" the checkbox can be found, labelled as "Keep Library Loaded - Check this box to retain the library in RAM". Once this box was enabled on the crashing server it immediately became stable. The particular CFX tag was the popular CFX_IMAGE. Having to load an unload the CFX into memory for every request was demonstrated to crash the server while leaving very little evidence in its wake.

Bones Festes i Bon Nadal!

This is my 4th time spending the holidays in Spain, one in Madrid and the last 3 in Barcelona, the heart of Catalonia where Spanish is the 2nd language between Catalan and English.

On the departing flight with Alitalia I managed to read the entire book "Teach yourself Unix in 10 minutes a day", an introduction to SELinux, an overview of the new features in Fedora Core 3, and a tutorial on building RPMs. Not bad overall, but I've got one hell of a jet lag from not sleeping.

[More]

Snap Gallery makeover for Belgium and Paris photos

Today I expanded my collection of online photos from my 1999 trip to Belgium for Christmas and Paris for New Years Eve on the Millenium. The gallery itself was rebuilt using the Snap Gallery software, so the photos now have titles and captions.

View Photo Gallery of Bruges, Brussels, Mechelen, and Paris.

Discussion on CFEXECUTE on Linux/Unix

Here are some excerpts from a discussion on CFTalk-Linux regarding memory utilization when using the CFEXECUTE tag on Linux/Unix. Many reported to have observed very high memory usage whlie using CFEXECUTE, and it suspected that ColdFusion (via the JVM and the System) calls fork().

This got me thinking, so I carried out some quick, but careful tests on Linux and Solaris while monitoring memory. I did not observe any significant impact on memory utilization, but that is not to say that the reports are wrong. The key is finding out what is different between the server behavior reported in the field and the server behavior observed in these tests.

I was wondering if anyone (Steve) could give me some more information about the performance of
> cfexecute on Linux with MX? I was told that the cfexecute tag is implemented using fork() and
> that each call can consume 100's of MB of memory? Is there any truth to this claim?


[More]

Previous Entries / More Entries