Time  Nick       Message
11:09 Elwell     Stupid Q - SIP2 in the koha sense isn't the same as SIP in the VoIP sense right?
11:09 Elwell     '''''''''''''';;''
11:10 Elwell     ah,found the 3M document
23:05 SelfishMan "my company emailaddresses on the  Bayesian spam filter black list."
23:05 SelfishMan oops, wrong channel
23:17 mason      for the NZ viewers -> http://wellingtonista.com/waitangi-weekend-venn-diagram
03:29 Amit       hi good morning koha
03:30 Amit       hi mason good morning
05:15 mason      hiya amit
05:15 Amit       hi mason
05:15 Amit       how is your weekend
05:16 mason      very good thanks
05:16 Amit       enjoying cricket NZ vs Aus
07:18 Elwell     heh. Love the venn diagram
07:18 mc         hello world
07:27 chris      hi mc
07:28 Elwell     Stupid Q - is it possible to use zebra for log parsing? kinda like splunk?
07:28 Elwell     (yes I know its not terribly koha related)
07:30 chris      you could
07:30 chris      there's a bunch of tools that would probably do a lot better job though
07:33 chris      with a lot less effort :)
07:35 Elwell     :-)
07:35 Elwell     wasnt sure how optimised zebra was
07:38 chris      its very fast at searching
07:40 chris      but you need to set up documents describing how your data is structured and what indices you want to build
07:41 chris      so you would need to set that up for you logfiles
07:42 chris      and querying your logfiles via Z39.50 or sru/sw just doesnt seem like a lot of fun :)
07:43 SelfishMan Although it is a rather geeky way to do it
07:44 SelfishMan Just don't Z39.50 enable a Furby to query your logfiles because that's just going too far
07:44 chris      no one should use Z39.50 unless they have to ;)
07:48 HARE9      good morning
07:51 chris      hi hare9
07:59 Amit       hi hare9
07:59 Amit       hi chirs have u check reading level in koha3
07:59 Amit       sory chris
08:02 chris      what do you mean?
08:03 Amit       i mean to say  http://lists.katipo.co.nz/public/koha/2006/010884.html
08:03 Amit       this one for koha 3
08:13 chris      i havent done it, but it might work, you would have to edit the zebra config
08:14 chris      hi paul
08:14 paul_p     hi chris
08:14 Amit       k
08:15 paul_p     & hi amit (& anyone else on this channel)
08:15 Amit       hi paul
08:16 kf         amit: perhaps this is the way to do it in 3.0: http://lists.katipo.co.nz/public/koha/2008-October/015763.html
08:17 Amit       k
08:18 Amit       thanks hope it will help
08:20 kf         did not try it yet, but i will need a new search option for our accession numbers
08:20 Amit       k
09:42 kf         hi there, i have a question: my colleague asked me, if its better to have a separate mysql-server. how do you handle that?
09:42 paul_p     is it better ? it just depends on your needs.
09:43 paul_p     if you have a dedicated mySQL server (with specific DB-tuning & backups & ...), then it's a good idea.
09:43 paul_p     the easiest way to do that is :
09:43 chris      and a fast network
09:43 paul_p     * install koha on a single server
09:43 chris      if your network is slow, then doing all the db work over the network is a big performance kill
09:44 paul_p     * modify the koha_conf.xml to link mysql from localhost to xxx.yyy.zzz.ttt
09:44 paul_p     or, if you install koha for the 1st time, you can, during configure, specify the address of the mySQL server
09:44 kf         util now, we have koha as vmware on one server, but we will probably host koha for more than one library
09:45 chris      you definitely wont want to run it under vmware in production
09:45 paul_p     chris: why ?
09:45 chris      emulation is always slower
09:46 kf         we do that for many applications atm, is the difference great?
09:47 kf         our koha seems ok for me, but the only person using it is me of course
09:48 kf         paul: do you use a separate server for your installations? or just a single server?
09:48 chris      paul: its also one more thing that can go wrong, and you end up supporting 2 OS's on the machine instead of one
09:49 chris      i can understand it for local applications
09:50 paul_p     chris: I agree with your ideas here. & we just use 1 server for our hosted koha
09:50 paul_p     in fact, we have 1 server, with vserver & 2 "vserver" machines : one for production one for testing
09:51 paul_p     vserser is not full virtualization.
09:51 chris      i usually do that with xen
09:51 chris      yeah we use vservers a lot at work
09:51 kf         ok, one more question. how to handle multiple koha installation?
09:51 kf         ah
09:52 paul_p     kf: we have developed some tools to help having multiple koha installations on the same server.
09:52 kf         vservers?
09:52 chris      you can do it the vserver, or xen etc
09:52 chris      or you can do it with multiple dbs and config files
09:52 paul_p     kf: we use vserver just for separating production & testing
09:52 chris      i currently have 6 koha's running on my server at home
09:53 paul_p     all our productions instances are on the same vserver.
09:53 chris      just with different databases and config files, but all using the same perl
09:53 paul_p     & the same Koha, chris is right
09:53 kf         i think i will need at least three - koha production, koha for training and one to test new versions
09:53 paul_p     then, it's different than what we have.
09:53 paul_p     for that, I suggest to have different VM
09:54 paul_p     just an example : I did an update at SAN-OP, and got a problem due to perl Text::CVS_XS, that has changed !
09:54 paul_p     (it use to encode files depending on locale, now it's latin1 if my investigations are correct)
09:55 paul_p     so, you need to have 2 or 3 completely differents environments
09:56 kf         what about local changes in templates? i will try to avoid it... but we will at least need a special search option for inventory numbers
09:57 kf         if you have just one koha for several installations its everywhere then, right?
09:58 kf         sorry, paul, i think i understand now
10:01 chris      i normally have a development, a staging and a production server
10:01 paul_p     chris: yes, and that can be virtual machines imo
10:02 chris      the production server could do both training and live
10:02 chris      (have 2 koha's on it)
10:02 paul_p     chris: ++
10:02 chris      staging is for testing
10:02 chris      and development is where you make your new templates etc
10:03 chris      often my staging box is a separate machine, and i run replication from the production box to it
10:04 kf         what about more than one production systems for different libraries?
10:04 chris      they would be on your production machine (or virtual machine)
10:05 chris      and that would be the machine you backup obsessively :)
10:05 chris      nothing helps you sleep better than reliable backups :)
10:06 chris      another option
10:06 chris      is using amazon's ec2
10:07 chris      where you just spark up another instance (from your koha image) for each production system
10:08 chris      http://aws.amazon.com/ec2/
10:24 kf         back, had a support emergency here - you always help chris
10:24 kf         but im not sure if i understand it right
10:24 kf         ok, you dont recommend a separate mysql-server due to performance reasons, is this right?
10:25 chris      it depends on your network
10:25 kf         and its possible to have more than one koha installation on one server, with different conf-files?
10:25 chris      if its 100
10:25 chris      Mbit (or better)
10:25 kf         ok
10:25 chris      then a machine tuned just to be a database server can help with performance
10:26 chris      it its slower, then the performance gain can be offset by network lag
10:27 chris      and yes you can have multiple koha's running on the same server
10:28 kf         we just start thinking about the best infrastructure for the production system
10:29 chris      ive done it both ways before (separate mysql machine and with mysql on the same machine)
10:30 chris      generally a webserver and a database server will compete for the same scarce resources (ram and cpu)
10:30 paul_p     kf: how many items in your catalogue ?
10:30 chris      so if its high usage, seperating them can be a good idea
10:30 chris      the nice thing is
10:30 chris      you can always start with them on the same server
10:31 chris      then shift it and change the config et voila ;)
10:31 chris      :)
10:31 kf         about 50.000 for our first library
10:32 kf         but some other libraries are also interested, so we have to think about more than one production system
10:33 kf         chris: your french is better than mine ;)
10:33 paul_p     kf: with 50 000 items, you don't need to care of having 2 separate servers imo !
10:33 chris      paul has heard me try to speak it, he may disagree :)
10:33 chris      yes, with 50,000 one will be fine
10:35 kf         but i think it would be easier to backup one db-server, instad of several mysql-instances in different vms?
10:35 kf         the question is about performance then i think, but that should not be a big problem when the network is fast enough
10:36 chris      ok, i should go to sleep
10:37 paul_p     sweet dreams chris
10:39 kf         sweet dreams chris
10:39 kf         and thx
10:40 kf         and thx paul, i will have to discuss this with my colleague, wo is responsible for the servers
10:41 kf         perhaps i will send him here next time :)