Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!newstop!texsun!convex!convex.convex.com!thurlow From: thurlow@convex.com (Robert Thurlow) Newsgroups: comp.protocols.nfs Subject: Re: NFS biod question Message-ID: Date: 5 Dec 90 22:05:04 GMT References: <565@bcstec.boeing.com> Sender: usenet@convex.com Lines: 29 In <565@bcstec.boeing.com> ced@bcstec.uucp (Charles Derykus) writes: >I have a question regarding biods and their use in NFS. >How does a biod function when transmitting data from a client to an NFS >server? The biod enhances performance by supporting read-ahead and write-behind. When you post write requests, the biod waits for the confirmation from the server on your behalf, and it tries to get a few server blocks into the buffer cache before they are needed. The system should perform exactly the same without the biods running, except that your processes will be slower at doing I/O over NFS. >The particular brand of NFS running on the server only allows a >cumulative total of 8K block i/o size per mount. >What specifically is the 8K limit? "cumulative total of 8K block i/o size per mount" doesn't parse for me. There is an 8K limit on the rsize/wsize transfer sizes (see mount(2)) which is caused by an underlying limit on the UDP packet size of the UDP/IP implementation in Sun and Sun-based architectures. I hope this helps, Rob T -- Rob Thurlow, thurlow@convex.com or thurlow%convex.com@uxc.cso.uiuc.edu ---------------------------------------------------------------------- "This opinion was the only one available; I got here kind of late." Brought to you by Super Global Mega Corp .com