If you are interested in trying out open source ILS options our migration may be of interest to you. We signed a contract with LibLime to migrate to Koha this year. The contract was for bibliographic record mapping, training and support for three years. Our users and circ records are migrating too. We chose Koha because we didn’t have to rebarcode our collection and because Koha will allow us to get our records in an XML format which makes converting them easier into other metadata schema. We control our records, and that is a good thing.
Right now we are getting ready to migrate. We have weekly conference calls with our LibLime assistant and she helps keep us on track with the data collection and matching our old information and classification from Winnebago’s Spectrum 5 to Koha’s. So tomorrow we should have an idea if we are on track for a go live date of Sept. 1. I have a fourteen page check list to complete before we go live.
We have scheduled a webinar for the 28th of July for a two hour training for the staff.
Our server is sitting outside my office door. We cannibalized an old Gateway server and upped the ram to 1 Gig. It will do fine for us to start. We also will have a server at a remote location to serve as a virtual server. It will shadow what we are doing on the library server, it will allow us to expand in the future and it also acts as a back up server.
These are the recommended configurations from a Koha comrade for our server for those of you who like this kinda stuff.
Recommended Hardware specifications for Butte Public Library
Xeon / Opteron Dual Core
4 Gb ECC [6+G preferred] , 8-16 G will give considerable performance boost.
2 * 80 Gb SATA [4 * 40GB preferred], formatted as XFS
RAID 1 (Mirrored disks) — [ Hardware RAID ]
Dual GB NICs.
Planning for system growth:
The single server configuration above is suitable for the collection and circulation requirements of BPL, giving moderate and stable performance.
Koha’s system requirements can be expected to outgrow this configuration as the number of supported libraries increases.
At this stage, a two- or three-server environment must be implemented, splitting the Zebra indexing engine and/or MySQL RDBMS services from the application + web server.
These services can be moved to new servers without affecting the core application/web server, thus allowing a relatively seamless hardware upgrade process, continuing to utilize the original server that supported the entire application.
Koha is comprised of three distinct services, each with its own resource usage profile.
Loosely, they are:
Web + Application services , cpu-intensive.
RDBMS (MySQL) service, memory and disk intensive.
Zebra indexing service, disk intensive.
Well that’s it for this entry- I’ll be back tomorrow with a progress report.