From owner-freebsd-questions Wed Jul 22 07:24:32 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA07698 for freebsd-questions-outgoing; Wed, 22 Jul 1998 07:24:32 -0700 (PDT) (envelope-from owner-freebsd-questions@FreeBSD.ORG) Received: from steffi.geodesic.com (steffi.geodesic.com [208.208.142.101]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA07643 for ; Wed, 22 Jul 1998 07:24:21 -0700 (PDT) (envelope-from rb@geodesic.com) Received: from [208.208.142.213] ([208.208.142.213]) by steffi.geodesic.com (8.8.8/8.8.8) with ESMTP id JAA00398; Wed, 22 Jul 1998 09:23:29 -0500 (CDT) (envelope-from rb@geodesic.com) X-Sender: rb@pop3.geodesic.com Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Wed, 22 Jul 1998 09:23:03 -0500 To: freebsd-questions@FreeBSD.ORG From: Richard Brooksby Subject: Problems with DPT SCSI controller Cc: Geodesic System Administrators Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG [Please retain Cc line on replies -- we are not subscribed to this list.] We have a FreeBSD 2.2.6 box which we are using for our mission-critical information system server. We decided that we needed a RAID for this machine to reduce the probability of failures which cost a lot of company time. We have a DPT SmartRAID IV Ultra. It appears to work OK, but when we started transferring our data to the RAID we got kernel messages warning of DPT errors and "salvaging ... from the jaws of destruction". I've attached the relevant "dmesg" output below. Our kernel configuration contains: controller dpt0 and options DPTOPT # will go away soon options DPT_MEASURE_PERFORMANCE # just enable for now options DPT_VERIFY_HINTR # Some hardware must have it! options DPT_TRACK_CCB_STATES # Some hardware must have it! options DPT_HANDLE_TIMEOUTS # Some hardware must have it! options DPT_TIMEOUT_FACTOR=4 # Some hardware needs more We've no idea what these do but the comments seemed to imply that switching them on was a good idea, so we did. Are these errors dangerous? What's causing them? Is it a configuration problem or hardware? How can we reconfigure things to make the RAID reliable? Thanks. --- dpt0 rev 2 int a irq 5 on pci1:7:0 dpt0: DPT type 3, model PM3334UW firmware 07L0, Protocol 0 on port 7410 with 458753MB Write-Back cache. LED = 0000 0000 dpt0: Enabled Options: Verify Lost Transactions Precisely Track State Transitions Collect Metrics Handle Timeouts dpt0 waiting for scsi devices to settle (dpt0:0:0): "DPT STEFFI 07L0" type 0 fixed SCSI 2 sd1(dpt0:0:0): Direct-Access sd1 could not mode sense (4). Using ficticious geometry 8600MB (17612800 512 byte sectors) dpt0 waiting for scsi devices to settle [...] dpt0 ERROR: Marking 153126 (Write (10) [6.1.18]) on c0b0t0u0 as late after 19280233usec dpt0: Salvaging Tx 153126 from the jaws of destruction (10000/23860005) dpt0 ERROR: Marking 543201 (Write (10) [6.1.18]) on c0b0t0u0 as late after 18033230usec dpt0 ERROR: Marking 543202 (Write (10) [6.1.18]) on c0b0t0u0 as late after 18034163usec dpt0 ERROR: Marking 543203 (Write (10) [6.1.18]) on c0b0t0u0 as late after 18033132usec dpt0 ERROR: Marking 543204 (Write (10) [6.1.18]) on c0b0t0u0 as late after 10499053usec dpt0 ERROR: Marking 543205 (Write (10) [6.1.18]) on c0b0t0u0 as late after 10499995usec dpt0 ERROR: Marking 543206 (Write (10) [6.1.18]) on c0b0t0u0 as late after 10500881usec dpt0 ERROR: Marking 543207 (Write (10) [6.1.18]) on c0b0t0u0 as late after 10501761usec dpt0 ERROR: Marking 543208 (Write (10) [6.1.18]) on c0b0t0u0 as late after 10502645usec dpt0 ERROR: Marking 543209 (Write (10) [6.1.18]) on c0b0t0u0 as late after 10503587usec dpt0: Salvaging Tx 543206 from the jaws of destruction (10000/17320375) dpt0: Salvaging Tx 543201 from the jaws of destruction (10000/24859218) dpt0: Salvaging Tx 543202 from the jaws of destruction (10000/24860581) dpt0: Salvaging Tx 543203 from the jaws of destruction (10000/24859224) dpt0: Salvaging Tx 543205 from the jaws of destruction (10000/17324756) dpt0: Salvaging Tx 543204 from the jaws of destruction (10000/17326172) dpt0: Salvaging Tx 543209 from the jaws of destruction (10000/17332051) dpt0: Salvaging Tx 543208 from the jaws of destruction (10000/17336467) dpt0: Salvaging Tx 543207 from the jaws of destruction (10000/17388941) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message