From owner-freebsd-questions Mon Feb 19 15:03:52 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id PAA01563 for questions-outgoing; Mon, 19 Feb 1996 15:03:52 -0800 (PST) Received: from cola.tiac.net (cola.tiac.net [199.0.65.55]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id PAA01558 for ; Mon, 19 Feb 1996 15:03:49 -0800 (PST) Received: from winkpc.zipnet.net ([204.176.87.131]) by cola.tiac.net (8.6.9/8.6.6.Beta9) with SMTP id SAA01133; Mon, 19 Feb 1996 18:03:01 -0500 Date: Mon, 19 Feb 1996 18:03:01 -0500 Message-Id: <199602192303.SAA01133@cola.tiac.net> X-Sender: ejon@mail.tiac.net X-Mailer: Windows Eudora Version 1.4.4 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: Terry Lambert From: ejon@tiac.net (Eric Jones) Subject: Re: NQNFS problems Cc: questions@freebsd.org Sender: owner-questions@freebsd.org Precedence: bulk > >The leases are revoked on reference; not before. Local cache coherency >is not maintained. I'm a little unclear on what you mean by "on reference" Do you mean on a write attempt? It seems to me that if a process on the server machine wants to write to a file which has leases against it, it ought to revoke those leases immediately, or at least when a client attempts to read additional data. > >What kind of clients are you using? > Pentium 133's with 64-128 MB of RAM (if you're wondering about the size of buffer caches). We're setting up a new ISP with the biggest FBSD cluster _I_'ve ever heard of. Unfortunately, we're having a VM problem/crisis (another message on that will follow on the questions list) which may monkeywrench the plans. >Caching NFS is evil. > Yeah. Well, it would be somewhat helpful on our read-only fairly static file systems (like /usr/local), but perhaps the price is too high stability- wise. Eric