From: Andreas U. <un...@ne...> - 2005-08-28 06:52:52
|
Hello, I have two disks - connected to a 3ware raid controller - which make a RAID1 array. smartd alerts, that one of these disks has 2 bad sectors. Since this, the raid controller sometimes removes the affected disk suddenly, without saying what's the exactly problem (I guess write erros to this sectors) from the array. So I want to give the badblockhowto a try. I removed the problematic disk from the raid1 array, and make it visible as a single scsi disks to linux. Also I disabled 3ware's write cache for this disk. I removed all partitions and create one big partition and formated it as ext3 filesystem. Then I have done all the steps written in the badblockhowto.txt. After invoking the selftest on this disk (and waiting for finishing), the counters for Offline_Uncorrectable and Current_Pending_Sector are still the same and haven't changed. Anyone has experince with this? Probably the controller redirects my write requests transparently to other sectors, when it get write errors from my requested sectors? I can imaging this in raid mode, but when it is a single disk... Cheers, Andreas |
From: Bruce A. <ba...@gr...> - 2005-09-07 14:54:51
|
Hi Andreas, There are some disks that don't reset Current_Pending_Sector to 0, even when it's readable again, or reallocated. I don't know why. Cheers, Bruce On Sun, 28 Aug 2005, Andreas Unterkircher wrote: > Hello, > > I have two disks - connected to a 3ware raid controller - which make a > RAID1 array. smartd alerts, that > one of these disks has 2 bad sectors. Since this, the raid controller > sometimes removes the affected disk > suddenly, without saying what's the exactly problem (I guess write erros > to this sectors) from the array. > > So I want to give the badblockhowto a try. I removed the problematic > disk from the raid1 array, and > make it visible as a single scsi disks to linux. Also I disabled 3ware's > write cache for this disk. I > removed all partitions and create one big partition and formated it as > ext3 filesystem. > > Then I have done all the steps written in the badblockhowto.txt. After > invoking the selftest on this disk > (and waiting for finishing), the counters for Offline_Uncorrectable and > Current_Pending_Sector are still > the same and haven't changed. > > Anyone has experince with this? Probably the controller redirects my > write requests transparently to > other sectors, when it get write errors from my requested sectors? I > can imaging this in raid mode, > but when it is a single disk... > > Cheers, > Andreas > > > > > ------------------------------------------------------- > SF.Net email is Sponsored by the Better Software Conference & EXPO > September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices > Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA > Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf > _______________________________________________ > Smartmontools-support mailing list > Sma...@li... > https://lists.sourceforge.net/lists/listinfo/smartmontools-support > > |
From: Andreas U. <un...@ne...> - 2005-09-07 18:12:48
|
Hi Bruce, I guess you are right, because since I made this procedure the disk hasn't been kicked out of the array. Perhaps I will try a firmware update on the Samsung SP1604N disk. Thx & Cheers, Andreas Bruce Allen wrote: >Hi Andreas, > >There are some disks that don't reset Current_Pending_Sector to 0, even >when it's readable again, or reallocated. I don't know why. > >Cheers, > Bruce > >On Sun, 28 Aug 2005, Andreas Unterkircher wrote: > > > >>Hello, >> >>I have two disks - connected to a 3ware raid controller - which make a >>RAID1 array. smartd alerts, that >>one of these disks has 2 bad sectors. Since this, the raid controller >>sometimes removes the affected disk >>suddenly, without saying what's the exactly problem (I guess write erros >>to this sectors) from the array. >> >>So I want to give the badblockhowto a try. I removed the problematic >>disk from the raid1 array, and >>make it visible as a single scsi disks to linux. Also I disabled 3ware's >>write cache for this disk. I >>removed all partitions and create one big partition and formated it as >>ext3 filesystem. >> >>Then I have done all the steps written in the badblockhowto.txt. After >>invoking the selftest on this disk >>(and waiting for finishing), the counters for Offline_Uncorrectable and >>Current_Pending_Sector are still >>the same and haven't changed. >> >>Anyone has experince with this? Probably the controller redirects my >>write requests transparently to >>other sectors, when it get write errors from my requested sectors? I >>can imaging this in raid mode, >>but when it is a single disk... >> >>Cheers, >>Andreas >> >> >> >> >>------------------------------------------------------- >>SF.Net email is Sponsored by the Better Software Conference & EXPO >>September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices >>Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA >>Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf >>_______________________________________________ >>Smartmontools-support mailing list >>Sma...@li... >>https://lists.sourceforge.net/lists/listinfo/smartmontools-support >> >> >> >> > > > >------------------------------------------------------- >SF.Net email is Sponsored by the Better Software Conference & EXPO >September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices >Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA >Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf >_______________________________________________ >Smartmontools-support mailing list >Sma...@li... >https://lists.sourceforge.net/lists/listinfo/smartmontools-support > > |
From: Bruce A. <ba...@gr...> - 2005-09-07 18:24:51
|
> I guess you are right, because since I made this procedure the disk > hasn't been kicked out of the array. Perhaps I will try a firmware > update on the Samsung SP1604N disk. I am pretty sure that the firmware update will zero all counters. Let us know if that works. Cheers, Bruce > Bruce Allen wrote: > > >Hi Andreas, > > > >There are some disks that don't reset Current_Pending_Sector to 0, even > >when it's readable again, or reallocated. I don't know why. > > > >Cheers, > > Bruce > > > >On Sun, 28 Aug 2005, Andreas Unterkircher wrote: > > > > > > > >>Hello, > >> > >>I have two disks - connected to a 3ware raid controller - which make a > >>RAID1 array. smartd alerts, that > >>one of these disks has 2 bad sectors. Since this, the raid controller > >>sometimes removes the affected disk > >>suddenly, without saying what's the exactly problem (I guess write erros > >>to this sectors) from the array. > >> > >>So I want to give the badblockhowto a try. I removed the problematic > >>disk from the raid1 array, and > >>make it visible as a single scsi disks to linux. Also I disabled 3ware's > >>write cache for this disk. I > >>removed all partitions and create one big partition and formated it as > >>ext3 filesystem. > >> > >>Then I have done all the steps written in the badblockhowto.txt. After > >>invoking the selftest on this disk > >>(and waiting for finishing), the counters for Offline_Uncorrectable and > >>Current_Pending_Sector are still > >>the same and haven't changed. > >> > >>Anyone has experince with this? Probably the controller redirects my > >>write requests transparently to > >>other sectors, when it get write errors from my requested sectors? I > >>can imaging this in raid mode, > >>but when it is a single disk... > >> > >>Cheers, > >>Andreas > >> > >> > >> > >> > >>------------------------------------------------------- > >>SF.Net email is Sponsored by the Better Software Conference & EXPO > >>September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices > >>Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA > >>Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf > >>_______________________________________________ > >>Smartmontools-support mailing list > >>Sma...@li... > >>https://lists.sourceforge.net/lists/listinfo/smartmontools-support > >> > >> > >> > >> > > > > > > > >------------------------------------------------------- > >SF.Net email is Sponsored by the Better Software Conference & EXPO > >September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices > >Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA > >Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf > >_______________________________________________ > >Smartmontools-support mailing list > >Sma...@li... > >https://lists.sourceforge.net/lists/listinfo/smartmontools-support > > > > > > > > ------------------------------------------------------- > SF.Net email is Sponsored by the Better Software Conference & EXPO > September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices > Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA > Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf > _______________________________________________ > Smartmontools-support mailing list > Sma...@li... > https://lists.sourceforge.net/lists/listinfo/smartmontools-support > > |
From: Bernd M. <melchers@ZEDAT.FU-Berlin.DE> - 2005-09-07 20:56:53
|
I have a Western Digital Disk (Model No WD800JB-00CRA1, Fw. 17.07W17, ~30 month old) which behaves strange in some ways. Sometimes there is an uncorrectable Sector at the same place which never gets reallocated by extended tests and never was detected by the western digital tools. The last item could lead to a problem when i try to replace the disk by warranty. As there are newer disks with newer firmware, i first want to give a better firmware a chance. But this raises a more general question: Are there any websites where one can download tools for extracting and burning hard disk firmware? And firmware itself? There is definitely a need for this, e.g. it is spoken that all disks within a raid array should behave best, when all of them have the same firmware. Is it necessary that all vendors have there own utilities or is there a common interface for updating firmware? > > hasn't been kicked out of the array. Perhaps I will try a firmware > > update on the Samsung SP1604N disk. > mit freundlichen Grüßen Bernd Melchers |
From: Andreas U. <un...@ne...> - 2005-09-08 04:02:57
|
You will not always get new firmwares for your disks. But sometimes you=20 get them from WD from their website (http://support.wdc.com/download/index.asp) or=20 contact the support. Andreas Bernd Melchers wrote: >I have a Western Digital Disk (Model No WD800JB-00CRA1, Fw. 17.07W17, >~30 month old) which behaves strange in some ways. Sometimes there >is an uncorrectable Sector at the same place which never gets >reallocated by extended tests and never was detected by the >western digital tools. The last item could lead to a problem >when i try to replace the disk by warranty. > >As there are newer disks with newer firmware, i first want to >give a better firmware a chance. But this raises a more general >question: Are there any websites where one can download tools >for extracting and burning hard disk firmware? And firmware itself? >There is definitely a need for this, e.g. it is spoken that all disks >within a raid array should behave best, when all of them have the same >firmware. Is it necessary that all vendors have there own utilities >or is there a common interface for updating firmware? > > =20 > >>>hasn't been kicked out of the array. Perhaps I will try a firmware >>>update on the Samsung SP1604N disk. >>> =20 >>> > >mit freundlichen Gr=FC=DFen >Bernd Melchers > > >------------------------------------------------------- >SF.Net email is Sponsored by the Better Software Conference & EXPO >September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practi= ces >Agile & Plan-Driven Development * Managing Projects & Teams * Testing & = QA >Security * Process Improvement & Measurement * http://www.sqe.com/bsce5s= f >_______________________________________________ >Smartmontools-support mailing list >Sma...@li... >https://lists.sourceforge.net/lists/listinfo/smartmontools-support > =20 > |