On 5/3/2010 11:35 PM, Andrew Deason wrote: > On Mon, 3 May 2010 23:04:06 -0400 > "Brandon S. Allbery KF8NH" <allb...@ece.cmu.edu> wrote: > >> On May 3, 2010, at 09:09 , Frank Burkhardt wrote: >>> is the openafs-fileserver supposed to take advantage of multiple >>> cpu cores? > > Yes, however, I think most of the cpu-intensive operations are only in > the RX layer (the network transport / RPC layer). And a lot of the RX > code is very serialized, AFAIR. So, you see a lot of usage on just one > core. > > It's one of the unfortunate shortcomings of the RX code. Is this > improved at all in 1.5-specific stuff? I don't remember.
The current rx statistics gathering code in Rx requires frequent access to a global lock that slows things down a lot. If statistics gathering is turned off (its on by default), then things do go faster. The serialization is per call and should not be blocking the receipt or transmission of data on multiple cores. The file server layer itself still shares global locks and those are more likely to be preventing multi-core scalability. Jeffrey Altman
smime.p7s
Description: S/MIME Cryptographic Signature
