[Linux-cluster] ext3 filesystem

[Matthew Geier]

Graham Wood wrote:
> 
>> That's what I was trying to say in the "they should all" quote above.
>> If it's ext3, it has to be mounted read-only everywhere because changes
>> to the file system will pull the
>> rug out from under ext3 on the other nodes, with unpredictable results.
>> If it's GFS, no problem sharing the file system.
> Sorry - the first post seemed to state that you could have one writing 
> and the rest reading, and I misread the second so missed that.

  I would think that the suggestion of one writer and multiple readers 
is NOT a good one. If all were Read Only, I guess it could work, but the 
moment one host is writing, I would think thats a sure fire way to cause 
a kernel panic in the other nodes when they go to read the ext3 file 
system and find it an inconsistent state as the writer is half way 
through some operation and the ext3 code never expected this state to 
occur on a read. And this is ignoring cache inconsistencies that could 
give incorrect or corrupt data.