Hi, working with a MySQL-Database (or s.th. similar) gives you always a single point of failure. Does anyone of you know a solution how to keep e.g. 2 MySQL datbases in sync all the time, so if one of them fails, no user of the db has to care of his data? Thanks for any hint to solve this. Ralf * * Ralf 'coko' Koch * mailto:info@formel4.de * --- Gute Nachricht: Windows 2000 ist 100% kompatibel. Die weltweit besten Forscher suchen nur noch, womit.
On Tuesday 15 January 2002 09:47, you wrote:
Hi,
working with a MySQL-Database (or s.th. similar) gives you always a single point of failure. Does anyone of you know a solution how to keep e.g. 2 MySQL datbases in sync all the time, so if one of them fails, no user of the db has to care of his data?
Thanks for any hint to solve this.
Ralf
*
Hi, the only way to have an "online" redundancy of data is trhough a wrapper. It may be provided for the guys of MySQL (i don't know if they have something like that) or you. The wrapper will listen transactions and then assign each transactions to both servers, so they can process and record it. Note that this kind of thing is very similar to journaling filesystems or raid (with obviuos diferences). The other form (not "online") is to keep the db files on rsync or cvs or such a thing and let them to take care of the data syncronization. Obviuosly there is still a chance to get a loss of data or data offset between databases. Bye -- Carlos Manuel Duclos Vergara
Hi Carlos, I'm looking only for an "online" redundancy - the only way not to loose information if one of the systems goes down. I'm not good in programming - scripting is the highest level I'm able of - so I need a some more hints how to build a working wrapper for this. Any idea? Thx, Ralf
On Tuesday 15 January 2002 09:47, you wrote:
Hi,
working with a MySQL-Database (or s.th. similar) gives you always a single point of failure. Does anyone of you know a solution how to keep e.g. 2 MySQL datbases in sync all the time, so if one of them fails, no user of the db has to care of his data?
Thanks for any hint to solve this.
Ralf
*
Hi, the only way to have an "online" redundancy of data is trhough a wrapper. It may be provided for the guys of MySQL (i don't know if they have something like that) or you. The wrapper will listen transactions and then assign each transactions to both servers, so they can process and record it. Note that this kind of thing is very similar to journaling filesystems or raid (with obviuos diferences).
The other form (not "online") is to keep the db files on rsync or cvs or such a thing and let them to take care of the data syncronization. Obviuosly there is still a chance to get a loss of data or data offset between databases.
Bye
-- Carlos Manuel Duclos Vergara
* * Ralf 'coko' Koch * mailto:info@formel4.de * --- Hiroshima 45, Tchernobyl 86, Windows 2000
On Tuesday 15 January 2002 10:18, Ralf Koch wrote:
Hi Carlos,
I'm looking only for an "online" redundancy - the only way not to loose information if one of the systems goes down. I'm not good in programming - scripting is the highest level I'm able of - so I need a some more hints how to build a working wrapper for this. Any idea?
Thx,
Ralf
Hi, look this web page: http://prometheus.zerodivide.net/mysql_cluster/ also in http://freshmeat.net there is a lot of good stuff of "toys" bye and good look -- Carlos Manuel Duclos Vergara
participants (2)
-
Carlos Manuel Duclos Vergara -
Ralf Koch