Tom Keiser wrote:
On 4/6/08, Jeffrey Altman <[EMAIL PROTECTED]> wrote:
Matt Benjamin wrote:

File Server/Cache Manager Extended Status Information

[snip]
2. the client has registered an interest in receiving extended
notifications

Proc RangeInvalidateCallBack

This RPC communicates a vector of file changes, of byte-range
granularity, for a corresponding vector of CallBacks.

proc RangeInvalidateCallBack {
~  IN  AFSCBFids *Fids_Array,
~  AFSRangeCallbackSeq *CallBacks_Array
} multi = 65539;


 Why is there a range of byte range changes?  A StoreData operation
 does not take a list of independent commits.  Each StoreData will
 result in a BreakCallBack().


Coalescing several quasi-simultaneous stores into a single callback
break rpc sounds like a natural optimization to me.  Furthermore, this
would simplify any future implementation of a vectorized store rpc,
compound rpcs (a la nfs4), etc.

-Tom
Sure but this proposal does not describe this as future proofing or how the existing
file server might be modified to perform the coalescing in a safe manner.

And if there is going to be coalescing then there needs to be support for coalescing not just the DataStore operations but also all of the other operations that can result in
callbacks being sent on the object.

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

_______________________________________________
AFS3-standardization mailing list
[email protected]
http://michigan-openafs-lists.central.org/mailman/listinfo/afs3-standardization

Reply via email to