Once again, there's been a fair amount of blogging activity among MarkLogic developers since the last update.
Alex Milowski provides more updates on his "Big Weather Data" saga (see previous installments here, here, and here), digging into the nitty-gritty nuances of configuring limited hardware resources for scale and performance. Parts read like a suspense novel (I'm not kidding!). My favorite quote (which also made the rounds on Twitter) was the conclusion (SPOILER ALERT): "Right now, I'm just a happy MarkLogic camper."
- Disk space is important!
- Too Much Data and Too Little Memory
- Experiments with Big Weather Data in MarkLogic - Right-sizing and Indexing
Dave Erickson shows how to use the Kettle Spoon tool to quickly set up a repeatable ETL job that moves data out of PostGreSQL and into MarkLogic:
For system administrators, we've got a recipe for cleaning up log files:
- Monitoring MarkLogic (Robin Bramley)
As well as a tutorial on how to configure rsyslog to monitor MarkLogic on Linux so that you can automatically be notified by email of important MarkLogic system events:
- rsyslog and MarkLogic (Michael Blakeley)
Jake Trent looks at different approaches to dynamic sorting in FLWOR expressions:
Dave Cassel shows how you can enable wildcards in dates in search queries, so a user could search, for example, any day this month by typing "date:2012-07-??":
Cassel also explores the "directory creation" setting, and circumstances under which you want to disable the default "automatic" setting (hint: almost always, which by the way is the reason the default value is changing in the next release). Be sure also to see Michael Blakeley's earlier Directory Assistance article.
- Directory Creation setting (Dave Cassel)
Here Cassel explains how MarkLogic calculates search results relevance (tf/idf) and how you can tweak the scoring algorithm on a per-database basis:
Relatedly, Ryan Dew explores how the "synonym" option in word queries impacts the order of search results:
Dew also provides an update on his in-memory update library for XQuery:
There's always room for esoterica among MarkLogic die-hards. Dew experiments with "unordered" expressions, and Cassel digs into a URL rewriting conundrum:
- Finding Order With The Unordered Expression In MarkLogic (Ryan Dew)
- Redirecting responses from the rewriter in MarkLogic (Dave Cassel)
On the languages front, Kurt Cagle imagines what it would be like to compile CoffeeScript into XQuery:
That's all for this installment. If you know of any MarkLogic bloggers I've missed or if you're going to start blogging yourself, please let me know. See you next time!