NFSv4, anyone?

[Nikola Milutinovic]

> To summarize, the situation isn't that different from the 'normal'
> situation. After all, even a single instance of cyrus running on one
> machine contains multiple imapds (at least abt one per connection, on
> my current, non-clustered system, abt 1500 at a time). And they all
> access the same config dir, the same spool. They are permorming just the
> waltz you mentioned - with file locks. They don't communicate by other
> means than the file system. The master process only spawns them. The
> same works with 'independent' instances as well. Read the threads. There
> are at least two sites doing just the thing mentioned, having had it in
> production use for years.

Thank you Janne for this insight. Your line about master spawning multiple imapd instances shed a lot of light. I realize you are right. I stand corrected.

> To spare you some time in reading through the threads, here is a pointer
> to some relevant information:
>
> http://irbs.net/internet/info-cyrus/0607/0449.html

Thanks.

Now, one question that is a bit off topic, why is NFSv4 locking so bad? Couldn't they (Powers that Be) fix it, so it behaves like a normal (albeit, networked) file system?

Nix.




__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com