From owner-freebsd-scsi Sun Dec 12 5:15:44 1999 Delivered-To: freebsd-scsi@freebsd.org Received: from mail.omnilink.net (mail.omnilink.net [194.64.25.6]) by hub.freebsd.org (Postfix) with ESMTP id 6A70D14E49 for ; Sun, 12 Dec 1999 05:15:41 -0800 (PST) (envelope-from ob@omnilink.net) Received: from ntsrv2 (my.jav.net [212.255.14.194]) by mail.omnilink.net (8.9.3/8.9.3) with SMTP id OAA01806 for ; Sun, 12 Dec 1999 14:17:05 +0100 (CET) (envelope-from ob@omnilink.net) Message-ID: <00e001bf44a2$ecea6c00$c20effd4@jav.net> From: "Oliver Blasnik" To: Subject: Again: CRD-Raid-Controller and FreeBSD 3.x Date: Sun, 12 Dec 1999 14:15:05 +0100 X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2014.211 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2014.211 Sender: owner-freebsd-scsi@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi there, I do have several FreeBSD-Systems (3.1, 3.3) connected to external CRD-5440-Raid-Controllers and encountered massive Problems on one (new) machine this weekend :( All are working with on-board Adaptec 7890-Controllers, set up to UW 40 MHz, which is in specification to that 5440. Connection is done via Granite-Cables and Terminators, the Raid is the only device, I do NOT have any termination- or cabling-Problems on any of that 3 machines. Firmware of the 5440s is 1.9. But what happend... The server refused to work on after multiple "nice" messages in /var/log/messages, telling me: Dec 11 04:01:57 bigfoot /kernel: (da1:ahc0:0:0:1): SCB 0x47 - timed out while idle, LASTPHASE == 0x1, SEQADDR == 0x9 Dec 11 04:01:57 bigfoot /kernel: (da1:ahc0:0:0:1): Queuing a BDR SCB Dec 11 04:01:57 bigfoot /kernel: (da1:ahc0:0:0:1): Bus Device Reset Message Sent Dec 11 04:01:57 bigfoot /kernel: (da1:ahc0:0:0:1): no longer in timeout, status = 34b Dec 11 04:01:57 bigfoot /kernel: ahc0: Bus Device Reset on A:0. 2 SCBs aborted This server is our backup-machine (amanda) and had high scsi-load at this time. I think that there were many SCB's waiting to finish and that driver didn't wait long enough. After that crash I checked out the other Raid-Systems and found identical messages on all (which were still running)... As a work-around the only thing was to disable tagged-queuing on all systems (added that devices to cam_xpt.c and build a new kernel). None of these messages occured anymore. Anyone experienced something like this? How could I expand the timeout for tagged-queuing? Is there any workaround to use tagq (scsi is now extremely slow in comparision to yesterday...)? Regards, Oliver To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message