On Mon, Feb 09, 2004 at 11:13:39AM -0800, Federico Sacerdoti wrote: > Yes, this is the case. The problem is you dont know how many nodes you > will have. What if a large number are added? Everyone's RRD disk > requirements will go up.
I'll make that a system config paramater with a reasionable default. I may make the parameter a minimum with the actual number plus a configurable amount of extra space. Making all the major parameters into configuration knobs is pretty much free anyway. > For Rocks clusters, we allocate a Linux tmpfs with a maximum size of > 25% available memory. Only the "frontend" node (not a compute node) > keeps the RRD database. FreeBSD doesn't have a growable tmpfs like that so we need to specify the full size at startup. Since many people will be running with small clusters, I'd like to keep the size down in those cases. -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
pgpEynQFIJz5l.pgp
Description: PGP signature
