#mysql
17 October 2007
Total 30 pages. You are browsing page 5/30.
First :: Prev :: [...] [3] [4] [5] [6] [7] [...] :: Next :: Last
04:13 <****> also. will mysql cacheing cause issues with parts of db changed by external app? or will mysql detect the changes and recache them?
04:14 <****> kevev: you should not have 2 process changing the data files
04:14 <****> kevev: caching in mysql will break
04:15 <****> kevev: and possible of data corruption too
04:15 <****> hi peeps i am so confused about mysql.
04:15 <****> i have been trying for hours to get it to work the way i want it to do
04:15 <****> ok. that is what I thought. So I should just use the replication built into MYSQL?
04:16 <****> i am trying to search for a category exact wording, search for a date and file name *.txt
04:17 <****> kimseong: What I am doing is Virtual IP Management between 3 nodes running Apache and MYSQL. They all replicate Apache data and config files. I need to replicate databases.
04:17 <****> kevev: what do you use to replicate
04:18 <****> kimseong: If a web page tries to write to MYSQL database on server holding Virtual IP and the database goes down, will the data be lost/
04:18 <****> I use Glusterfs
04:18 <****> kevev: you probably better off with 1 single database
04:18 <****> I am guessing it depends on the webapps timeout.
04:18 <****> ok
04:19 <****> is there a nice website out there with examples of searching for stuff instide a database?
04:19 <****> inside a dataabse table?
04:20 <****> !m bbelt16ag select
04:20 <****> bbelt16ag: (SELECT Syntax) : http://dev.mysql.com/doc/mysql/en/select.html
04:20 <****> !m bbelt16ag fulltext
04:20 <****> bbelt16ag: (Full-Text Search Functions) : http://dev.mysql.com/doc/mysql/en/fulltext-search.html
04:21 <****> kimseong: well I want to have all 3 servers replicating in a round robin. I know how to implement this. I am just wondering if a webapp accessing the db being pointed to by the VIP will cause damage to the db on other 2 servers if server 1 goes down during a write.
04:21 <****> ok ok i just found those just this second and i switch windows..
04:22 <****> seesh
04:22 <****> this is complicated
04:22 <****> looks like my database is going to be wrapped in some bashscript to make it work better and easier for me.
04:23 <****> kevev: if 1 single db server, the webserver wither has send the write to the db or has not before it goes down
04:24 <****> kevev: if 3 db servers, thats is tricky, you can only do this with external-locking with myisam tables and 0 caches, the 0 caches is slow, might as well have 1 db server with caches enabled
04:24 <****> I need replication
04:25 <****> I need 0 points of failure.
04:25 <****> kevev: if you want to cover for the potential db failure, then have a backup server, there is a HA whitepaper on mysql website
04:25 <****> on the methods you can use
04:25 <****> ok
04:25 <****> 0 points of failure? usually people looks for 1 point of failure, seldom 2 points
04:26 <****> kimseong: OK I will not use external replication. I will MYSQL builtin replication.
04:26 <****> kevev: mysql replication is asynchronous, thats something to be aware of
04:27 <****> kimseong: If I do round robin replication through MYSQL replication will a write to one server during a crash cause issues? or will the data just not be replicated until the server is brought back online?
04:27 <****> ohh..asynchronous. I thought it was instant
04:28 <****> is there a way to make it instant?
04:29 <****> kevev: remove all latency and cpu limitations.
04:29 <****> oh, and add 1TB of RAM on each machine.
04:29 <****> damn.
04:30 <****> I do not want to run whole db from cache.
04:30 <****> (and an OS and architecture that can support 1TB ram)
04:30 <****> jpalmer: lol
04:30 <****> I run Linux ;o)
04:30 <****> can I do a shared cache between servers?
04:31 <****> kevev: use oracle rac
04:33 <****> kevev: marathon software has a product that share a virtual machine across 2 physical server, it is like shared process+all resources, but run on windows and it will be expensive
04:33 <****> Hi - how do i add a clause to return records based on date -- something date <= 10/16/07 ?
04:33 <****> ragas: put the date between quotes.
04:33 <****> i.e. '16-10-2007'
04:33 <****> kimseong: thanks. But we are sticking with linux. Need something free.
04:33 <****> ragas: is the data type a date type? or a string?
04:34 <****> kevev: read the whitepaper, you probably can find something free in there too
04:34 <****> i believe its a data type. I forget why i did that..
04:34 <****> a string, i mean
04:35 <****> ragas: you should always store dates as datetime
04:35 <****> if string, good luck, do your own comparision based on string functions
04:35 <****> i couls change it
04:35 <****> ^could
04:35 <****> you should change it.
04:36 <****> i think the formatting was a problem, getting it to store correctly. I was lazy, im sure.
04:36 <****> thank you all.
04:36 <****> bye
04:36 <****> thanks. then i can use the syntax i mentioned, with the single quotes, eh?
04:37 * chadmaynard would like 1TB of RAM
04:37 <****> DDR4 that is
04:37 <****> chadmaynard: I heard you had a good score?
04:37 <****> ragas: reliably, yes
04:37 <****> thumbs: not really
04:38 <****> chadmaynard: well I beat 7100
04:38 <****> cool
04:39 <****> DRAM? i want SRAM faster access time
04:40 <****> if I have a table that has a column called "weight." Is there a way to query the next record that has has a weight closest to n without going over?
04:40 <****> kimseong: DDR4 is planned to be faster than any sram in the works.
04:40 <****> and visa-versa (record closest to n without going under)
04:40 <****> people still haven't adapted DDR3 yet
04:40 <****> unless we're talking about the video game
04:40 <****> :-P
04:41 <****> dram probable better for sequential, sram should still be better for random
04:43 <****> any idea on the SQL question. I tried "SELECT * FROM photos WHERE weight < 5 LIMIT 1" but that just returns the first ...oh wait
04:43 <****> SELECT * FROM photos WHERE MAX(weight) < 5 LIMIT 1
04:43 <****> no wait..that won't work
04:43 <****> haha. it returns 1 because you said limit 1 genius.
04:45 <****> hmm..any ideas?
04:45 <****> Inv1s1ble: it returns 1 because you said limit 1
04:45 <****> max is aggregate function, not usable in WHERE
04:46 <****> Inv1s1ble: you probably wants ORDER BY
04:46 <****> unless i am missing your issue
04:47 <****> I'd need a nested query wouldn't I? so I could order first
04:48 <****> You know the " -- " is used to ignore the rest of the query in MS SQL does mysql have something similar?
04:48 <****> Inv1s1ble: you can use aggregates in the having clause. If you want things in the row related to the MAX you'll need a subquery
04:48 <****> viro: comment ?
04:48 <****> kimseong: as in // ?
04:48 <****> Inv1s1ble: you probably wants ORDER BY
04:48 <****> this wokrs: SELECT * FROM (SELECT * FROM photos ORDER BY weight DESC) AS t WHERE weight < 4 LIMIT 1
04:48 <****> !m viro comment
04:49 <****> viro: (Comment Syntax) : http://dev.mysql.com/doc/mysql/en/comments.html
04:49 <****> where 4 is the element I'm trying to find the previous to
Total 30 pages. You are browsing page 5/30.
First :: Prev :: [...] [3] [4] [5] [6] [7] [...] :: Next :: Last