A while ago, I was paged by a client having backup job failure. Doing my routine as usual, I started the investigation by looking at the job execution log. Connected to the box was a strange output. Here it is!
Sharing knowledge not only is beneficial to others, but also enhances one’s own knowledge and broadens the spectrum. Blogging is all about sharing and this Log Buffer Edition brings together some great blog posts from around the database arena.
Recently, I was upgrading a database from 22.214.171.124 to the current 126.96.36.199 version. The database was using ASM, but I should notify at the beginning that the configuration is for a Stand-Alone Server and not RAC. Basically, the first things to be done for this procedure are part of the following checklist…
Join Francisco Bordenave from Pythian’s MySQL team for a presentation on replication, old and new.
The traditional NoCOUG SQL challenge has been launched this year with a twist: In the wake of the “BigData” trend/buzz, it’s now been upgraded to a “SQL and/vs. NoSQL” challenge. I took on the challenge, treading through my SQL comfort zone and thinking of ways I could bend relational algebra to solve the wicked puzzle suggested this year.
What if we added a new field in the META.json — let’s call it x_help_wanted — that would contains all the different types of help a module maintainer could require? Positions like maintainer, co-maintainer, coder, translator, documentation, and tester. We could even have a Dist::Zilla plugin to populate that field for us.
This is going to be a short one, but I think the changes to GitStore are cool enough to deserve a little blog-squeal.
I played with Mongo and looked at Mongoose, which are nifty, but holy schmolee are Mongo databases huge. And then I re-discovered DBIx::NoSQL, which was pretty much smack what I wanted. But I needed a way to easily serialize my objects for it. So I dragged in MooseX::Storage to the mix. And then I had fun with helper classes and roles to make the interfacing between the two systems as smooth and slick as a buttered piglet.
With the rapid advancement in the database technologies, the legacy systems are either being upgraded or replaced. Or, in some cases, technologists are finding ways to support them in new ways, showing us the flexible nature of databases and the belief of professionals that the sky is the limit. For this Log Buffer Edition, we go even beyond.
A new trial version of DBD::Oracle has been churned out. This release is mostly about Martin J. Evans going all ninjawesome on minor bug fixes as well as paving the way for an upcoming refactoring/speed boost of ora_verbose. As usual, the new version will be soaked for at least 2 weeks before it will turn into its fit-for-general-consumption v1.46 incarnation. Testers, please give this baby a whirl. The full changelog follows for the curious-minded.