Re: mysqldump question

[Aaron Wohl]

http://www.birdsoft.demon.co.uk/proglib/slowpipe.htm would seem to do
what you want... I havent tried it yet, but noted the URL for the next
time I needed that functionality.


- Original message -
From: "Amit M Bhosle" <[EMAIL PROTECTED]>
To: mysql@lists.mysql.com
Date: Fri, 13 May 2005 09:18:00 +0530
Subject: mysqldump question

Hi:

 i was wondering if there's any way to limit the bandwidth used by
mysqldump to dump data from remote hosts. since i couldn't find any
documentation on this, i assume that mysqldump will use all the
available bandwidth of the network.

 the issue is that i'm looking to fetch data to the tune of 100s of
MBs, and i don't want the mysqldump to hog all the bandwidth, thus
adversely affecting other communication.

thx in advance for ur time.
AB

-- 
A great idea need not be complicated.
http://www.cs.ucsb.edu/~bhosle

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

Re: Rollback

[Aaron Wohl]

You can add a version field to each row.  Then add a seperate table with
info with a list of the versions and a flag for deleted.  Queries would
look for each record that has the highest version number thats not
deleted. Having a lot undo/redo info can get kind of complicated,
especialy with multiple end users playing with it and chains of dependant
changes.

If the info can be modeled as documents this is frequenetly done with
CVS.  

On Fri, 2 Jan 2004 13:06:36 +0530, "karthikeyan.balasubramanian"
<[EMAIL PROTECTED]> said:
> Hi,
> 
>   I posted this question in MySQL mailing list and got no reply.
> 
> The basic problem is that I have committed the transaction and then
> replicated to another DB. Now I want to rollback the committed
> transaction.
> Is there a way to rollback to a particular point. This requirement is
> very
> similar to rolling back using save points. I guess an option would be to
> backup database before changes and restore it if the user is not
> satisfied
> with the changes he has made. One transaction in my application would
> affect
> 6-8 tables with at least 50 - 100 records getting inserted/updated or
> deleted.
> 
> Please advice
> 
> PS : Wish you all a very Happy New Year
> 
> Karthikeyan B
> 
> 
> 
> 
> -- 
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
> 

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]

Re: Migration to INODB?

[Aaron Wohl]

Run all your tables thru this.  Save the script as xinno_convert.  As in
xinno_convert < foo.sql | mysql

Suposedly you can also alter a table to innodb.  But that never worked
for me it just looped using cpu for days.  So I made this script which
worked fine.  Other than converting implicit locking is different.  With
innodb it assumes your using transactions.


#!/usr/bin/perl

#convert all tables in a mysql dump to innodb
use strict;
while(<>) {
  my $aline=$_;
  if($aline=~/(\) TYPE=)[A-Za-z]*ISAM(.*)$/) {
$aline="$1INNODB;\n";
# ) TYPE=MyISAM COMMENT='Users and global privileges';
  }
  print $aline;
}

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]