[activegrid-developers] Replicated storage
Brought to you by:
activegrid,
geremycohen
From: Mike J. <ims...@gm...> - 2005-05-14 23:25:26
|
Hi, Heard about your software on ITConversations.com and decided to take a look for the project I'm currently scoping out. I've read through most of the architecture whitepaper and on lots of the diagrams in the 'Deployment Patterns' section the DB appears as if on another machine and some of the other diagrams seem to totally ignore the M in LAMP. What I'm interested in is how it would work if you wanted to follow the Inter-Node pattern but not seperate out the DB element onto another machine/s. So it wouldn't matter if you had 1 machine or 100 but that the data was replicated amongst them. Also on a file level, as I've always been taught its best to store files (images etc) in the file system and store pointers in DBs (this may not be such as issue any more but I've not heard any different). Would this need to be run in conjuction with MySQL replication (never used it so please be gentle with explainations!) and something like DRBD for filesystem replication (again never used it but heard of it). Looking forward to hearing your thoughts and experimenting with your stuff. Regards, Mike --=20 "Opportunities are usually disguised as hard work, so most people don't recognize them." - Ann Landers |