Date: Tue, 4 Nov 2003 14:51:15 +0100 From: Francesco Casadei <fcasadei@inwind.it> To: Dmitry Alyabyev <dimitry@al.org.ua> Cc: freebsd-isp@freebsd.org Subject: Re: ad7: removed from configuration Message-ID: <20031104135115.GA2289@goku.kasby> In-Reply-To: <200311041425.57134.dimitry@al.org.ua> References: <200311041425.57134.dimitry@al.org.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
--VS++wcV0S1rZb1Fb Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, Nov 04, 2003 at 02:25:57PM +0200, Dmitry Alyabyev wrote: > FreeBSD 4.6.1-RC2 > hw.ata.tags: 0 >=20 > Oct 24 14:05:32 <kern.crit> /kernel: atapci1: <Promise TX2 ATA100 control= ler> port 0xef90-0xef9f,0xefe0-0xefe3,0xefa8-0xefaf,0xefe4-0xefe7,0xeff0-0x= eff7 m > em 0xffae0000-0xffaeffff irq 9 at device 10.0 on pci0 > Oct 24 14:05:32 <kern.crit> /kernel: ata2: at 0xeff0 on atapci1 > Oct 24 14:05:32 <kern.crit> /kernel: ata3: at 0xefa8 on atapci1 > Oct 24 14:05:33 <kern.crit> /kernel: ad4: 95396MB <WDC WD1000BB-32CCB0> [= 193821/16/63] at ata2-master UDMA100 > Oct 24 14:05:33 <kern.crit> /kernel: ad5: 95396MB <WDC WD1000BB-00CAA0> [= 193821/16/63] at ata2-slave UDMA100 > Oct 24 14:05:33 <kern.crit> /kernel: ad6: 114473MB <WDC WD1200JB-00CRA1> = [232581/16/63] at ata3-master UDMA100 > Oct 24 14:05:33 <kern.crit> /kernel: ad7: 114473MB <WDC WD1200JB-00DUA3> = [232581/16/63] at ata3-slave UDMA100 >=20 > and all works just fine.=20 > BUT after long period of time I got: >=20 > Oct 26 00:20:52 <kern.crit> /kernel: ad7: READ command timeout tag=3D0 se= rv=3D0 - resetting > Oct 26 00:20:52 <kern.crit> /kernel: ata3: resetting devices .. > Oct 26 00:20:52 <kern.crit> /kernel: ad7: removed from configuration > Oct 26 00:20:52 <kern.crit> /kernel: done >=20 > ... and after several hours the server stoped working. Power off/on made = it works fine again >=20 > any ideas about source of the problem (I mean "removed from configuration= ") ? > any solution (ad7 is not system disk) ? >=20 > --=20 > Dimitry >=20 > _______________________________________________ > freebsd-isp@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-isp > To unsubscribe, send any mail to "freebsd-isp-unsubscribe@freebsd.org" > end of the original message Unfortunately, I don't know the answer to your question, but I want to share my experience because it looks very similar. I have a Promise TX2 ATA RAID controller (long output not wrapped): # pciconf -l -v [...] atapci1@pci0:9:0: class=3D0x010485 card=3D0x4d68105a chip=3D0x6268105a re= v=3D0x02 hdr=3D 0x00 vendor =3D 'Promise Technology Inc' device =3D 'PDC20268R FastTrak100 TX2/TX4/LP EIDE controller' class =3D mass storage subclass =3D RAID [...] Two IBM disks are attached to the controller, and another IBM disk is attached to the mainboard controller: # atacontrol list ATA channel 0: Master: ad0 <IC35L040AVER07-0/ER4OA44A> ATA/ATAPI rev 5 Slave: no device present ATA channel 1: Master: acd0 <LG CD-ROM CRD-8521B/1.03> ATA/ATAPI rev 0 Slave: no device present ATA channel 2: Master: ad4 <IC35L040AVER07-0/ER4OA44A> ATA/ATAPI rev 5 Slave: no device present ATA channel 3: Master: ad6 <IC35L040AVER07-0/ER4OA44A> ATA/ATAPI rev 5 Slave: no device present All three disks are identical. One bad day, I discovered that the master drive on the third channel disappeared. Swapping the power cable between the two drives ad4 and ad6 made the latter appear again, but now it is recognized as a ~33GB drive (it was a 40GB drive as the code IC35L040AVER07-0/ER4OA44A should prove). I ran IBM/Hitachi DFT (Drive Fitness Test) utility on ad6 and it says the drive is ok, even if the DFT program itself lists the drive as a ~33GB disk. Here's the output from atacontrol for the working drive: # atacontrol cap 2 0 ATA channel 2, Master, device ad4: ATA/ATAPI revision 5 device model IC35L040AVER07-0 serial number SX0SX058089 firmware revision ER4OA44A cylinders 16383 heads 16 sectors/track 63 lba supported 80418240 sectors lba48 not supported =20 dma supported overlap not supported Feature Support Enable Value Vendor write cache yes yes read ahead yes yes dma queued yes yes 31/1F SMART yes yes microcode download no no security yes yes power management yes yes advanced power management yes no 0/00 automatic acoustic management yes no 254/FE 128/80 and here's the output for the 'failing' drive: # atacontrol cap 3 0 ATA channel 3, Master, device ad6: ATA/ATAPI revision 5 device model IC35L040AVER07-0 serial number SX0SX054672 firmware revision ER4OA44A cylinders 16383 heads 16 sectors/track 63 lba supported 66055248 sectors lba48 not supported =20 dma supported overlap not supported Feature Support Enable Value Vendor write cache yes yes read ahead yes yes dma queued yes yes 31/1F SMART yes yes microcode download no no security yes yes power management yes yes advanced power management yes no 0/00 automatic acoustic management yes no 254/FE 128/80 Look at the reported LBA size: lba supported 80418240 sectors # ad4 lba supported 66055248 sectors # ad6, WRONG should be 80418240! Now I come to the conclusion: this problem, plus other problems I didn't mention, make me think that the Promise TX2 ATA RAID card is a VERY bad controller, maybe it's too cheap to just work! If you find a solution to your problem, that may also solve mine, then I'd really like to hear from you! Francesco Casadei --=20 You can download my public key from http://digilander.libero.it/fcasadei/ or retrieve it from a keyserver (pgpkeys.mit.edu, wwwkeys.pgp.net, ...) Key fingerprint is: 1671 9A23 ACB4 520A E7EE 00B0 7EC3 375F 164E B17B --VS++wcV0S1rZb1Fb Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (FreeBSD) iD8DBQE/p67TfsM3XxZOsXsRAjtpAJ0RHZbuDkuTAvLDFobtxmu5NNyPIQCg7wtP zUxihZg2X6sQqtXBst8ONMM= =t4y1 -----END PGP SIGNATURE----- --VS++wcV0S1rZb1Fb--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20031104135115.GA2289>