Re: Ouch! ibdata files deleted. Why no catastrophe?

2007-09-25 Thread Gary Josack
For future reference. The files do actually continue to be written to. I experience this all the time when people delete logs files and space keeps filling up. Daniel Kasak wrote: On Tue, 2007-09-25 at 23:11 -0400, Gary Josack wrote: Well if you can stop all instances of writes to the data

Re: Ouch! ibdata files deleted. Why no catastrophe?

2007-09-25 Thread Daniel Kasak
On Tue, 2007-09-25 at 23:11 -0400, Gary Josack wrote: > Well if you can stop all instances of writes to the databases you should > be able to recover them. > > Each file is going to be in /proc/5460/fd/10-17 > > the file number corresponds to the fd you see in lsof output > > ex: > cp /proc/54

Re: Ouch! ibdata files deleted. Why no catastrophe?

2007-09-25 Thread Gary Josack
Well if you can stop all instances of writes to the databases you should be able to recover them. Each file is going to be in /proc/5460/fd/10-17 the file number corresponds to the fd you see in lsof output ex: cp /proc/5460/fd/10 ibdata2 This is still risky and i reccomend you get a dump imm

Re: Ouch! ibdata files deleted. Why no catastrophe?

2007-09-25 Thread Daniel Kasak
On Tue, 2007-09-25 at 19:27 -0400, Gary Josack wrote: > Did the space become available when deleted? > > try: > lsof | grep deleted > > see if they're still running in memory. if so you might be able to save > them. Thanks for the quick response :) They're there: mysqld 5460 mysql

Re: regexp negate string help

2007-09-25 Thread Baron Schwartz
MySQL's regex library doesn't have all those Perl features. You can use the pcre-compatible extension from http://www.xcdsql.org/MySQL/UDF/, or just use two clauses in the WHERE: one should be col NOT RLIKE "linux$" Baron Tang, Jasmine wrote: Hi, I need to match anything that start with "

Re: Ouch! ibdata files deleted. Why no catastrophe?

2007-09-25 Thread Gary Josack
Did the space become available when deleted? try: lsof | grep deleted see if they're still running in memory. if so you might be able to save them. Daniel Kasak wrote: Greetings. I've just returned from holidays, and it seems that all but 1 ibdata file ( there were 10! ) have been deleted b

Ouch! ibdata files deleted. Why no catastrophe?

2007-09-25 Thread Daniel Kasak
Greetings. I've just returned from holidays, and it seems that all but 1 ibdata file ( there were 10! ) have been deleted by a co-worker. He apparently was able to delete them with nautilus ( he was looking to reclaim some space and these were 1GB files each ... and yes, the Trash was emptied as w

RE: Ordering by unrelated column in a GROUP BY

2007-09-25 Thread Rob Wultsch
Peter, Thank you for your reply. MAX(t1.occurrence ) will pull the max of the occurrence column out of the group, but the other collumns (like data3 or id) would still be sorted by the GROUP BY. I will try your second solution, but the tables I am working are thousands of row and your solution lo

Ugly sql optimization help?

2007-09-25 Thread Bryan Cantwell
I have the following horrible sql. I need one result that has all the data in one row. I am currently using 3 sub queries and figure it must be a better way... SELECT 'FS_DEV', ifnull(a.severity, 0) AS aseverity, ifnull(a.

Re: Ordering by unrelated column in a GROUP BY

2007-09-25 Thread Peter Brawley
You might like to compare the performance of ... SELECT t1.data1, t1.data2, MAX(t1.occurrence) FROM t1 GROUP BY data1,data1 ORDER BY occurrence; with... SELECT t1.data1, t1.data2,t1.occurrence FROM t1 LEFT JOIN t1 AS t2 ON t1.data1=t2.data2 AND t1.data2=t2.data2 AND t1.occurrence < t2.occurre

Ordering by unrelated column in a GROUP BY

2007-09-25 Thread Rob Wultsch
Suppose I have a table: CREATE TABLE `t1` ( `id` int(11) NOT NULL auto_increment, `data1` varchar(50) , `data2` varchar(50) , `data3` varchar(50) , `occurance` datetime , PRIMARY KEY (`id`) ) And I want to pull the most recent entry of each set of unique combinations of `data1` and `d

RE: csv to mysql

2007-09-25 Thread Jay Blanchard
[snip] : i,m looking for a solution for my PDA that Doesn't have a DB Solution installed on it : so im having to write to CSV Files for my Forms , i'm needing a way that when i sink my PDA with my wireless network it Moves the Entire CSV File into a MySQL database :any Suggestions :? [/sni

csv to mysql

2007-09-25 Thread Brian E Boothe
hi all : i,m looking for a solution for my PDA that Doesn't have a DB Solution installed on it : so im having to write to CSV Files for my Forms , i'm needing a way that when i sink my PDA with my wireless network it Moves the Entire CSV File into a MySQL database :any Suggestions :? -- M

Re: more elegant way to store/find phone numbers

2007-09-25 Thread Peter Brawley
> does anyone have a nicer solution for this? How about comparing ereg_replace( "[[:punct:]]","", $colvalue ) with ereg_replace( "[[:punct:]]","", $comparisonvalue )? PB mysql wrote: hi listers we have a mysql based application, wherein phone numbers may be stored and searched for. it is no

RE: more elegant way to store/find phone numbers

2007-09-25 Thread Edward Kay
> hi listers > we have a mysql based application, wherein phone numbers may be stored > and searched for. it is not the primary goal of this application to > handle phone numbers. > > phone numbers usually are entered in a form like 099 999 99 99 or > 099-999-99-99, or substings thereof. actually,

more elegant way to store/find phone numbers

2007-09-25 Thread mysql
hi listers we have a mysql based application, wherein phone numbers may be stored and searched for. it is not the primary goal of this application to handle phone numbers. phone numbers usually are entered in a form like 099 999 99 99 or 099-999-99-99, or substings thereof. actually, the appl