From owner-freebsd-bugs Fri Nov 8 12:14:41 1996 Return-Path: owner-bugs Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id MAA05948 for bugs-outgoing; Fri, 8 Nov 1996 12:14:41 -0800 (PST) Received: from Kitten.mcs.com (Kitten.mcs.com [192.160.127.90]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id MAA05941; Fri, 8 Nov 1996 12:14:37 -0800 (PST) Received: from Mailbox.mcs.com (Mailbox.mcs.com [192.160.127.87]) by Kitten.mcs.com (8.8.2/8.8.Beta.3) with ESMTP id OAA02803; Fri, 8 Nov 1996 14:14:34 -0600 (CST) Received: from Mars.mcs.net (mikebo@Mars.mcs.com [192.160.127.85]) by Mailbox.mcs.com (8.8.2/8.8.2) with ESMTP id OAA06140; Fri, 8 Nov 1996 14:14:32 -0600 (CST) Received: (from mikebo@localhost) by Mars.mcs.net (8.8.2/8.8.2) id OAA24913; Fri, 8 Nov 1996 14:14:30 -0600 (CST) From: Michael Borowiec Message-Id: <199611082014.OAA24913@Mars.mcs.net> Subject: /kernel: nfsd send error 55 (No buffer space available) To: hackers@freebsd.org Date: Fri, 8 Nov 1996 14:14:30 -0600 (CST) Cc: bugs@freebsd.org Reply-To: mikebo@tellabs.com X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-bugs@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I sent this message to the "questions" list, but got nothing... Greetings - I have a Toshiba Tecra 720CDT w/48MB and a 3Com 3C589C PCCARD ethernet. I'm configuring this machine as an NFS server with 4 nfsd's running... My client machine is a 486DX50 w/16MB and a WD8013EPC ISA ethernet. I'm configuring this machine as the NFS client with 4 nfsiod's running... Both are running 2.1.5R. After copying a few files from server to client, the transfer freezes, and the server (laptop) starts displaying this error message, approx. once every 30 seconds: server /kernel: nfsd send error 55 Of course, this is the dreaded mbuf problem, confirmed by trying a ping: server# ping client PING client (198.102.156.3): 56 data bytes ping: sendto: No buffer space available ping: wrote client 64 chars, ret=-1 ... I rebuilt the server's kernel with more mbuf's, thinking this might solve it: options "NMBCLUSTERS=2048" but get the same problem... and the client must be rebooted each time. The server can be restored to "sanity" by killing all nfsd's and down-ing, then up-ing the interface (zp0). But to play it safe, both client and server were rebooted - same problem. server# netstat -m 155 mbufs in use: 115 mbufs allocated to data 34 mbufs allocated to packet headers 4 mbufs allocated to protocol control blocks 2 mbufs allocated to socket names and addresses 34/34 mbuf clusters in use 87 Kbytes allocated to network (100% in use) 0 requests for memory denied 0 requests for memory delayed 0 calls to protocol drain routines It doesn't look like many clusters are allocated at all... Does anyone know what might be causing this? Any suggestions? - Mike