Re: [gentoo-user] RAID: new drive on aac raid
Am 07.10.20 um 10:40 schrieb Stefan G. Weichinger: > Am 06.10.20 um 15:08 schrieb k...@aspodata.se: >> Stefan G. Weichinger: >>> I know the model: ICP5165BR >> >> https://ask.adaptec.com/app/answers/detail/a_id/17414/~/support-for-sata-and-sas-disk-drives-with-a-size-of-2tb-or-greater >> >> says it is supported up to 8TB drives using firmware v5.2.0 Build 17343 ** >> >> ** Firmware v5.2.0 Build 17343 for the ICP5045BL, ICP5085BL, ICP5805BL, >>ICP5125BR, and ICP5165BR: Adaptec is providing minimally tested >>firmware packages. Please contact Adaptec by PMC Technical Support >>to obtain these firmware files. Have the TSID or serial number of >>the product at hand when contacting support. > > Yes, I saw that as well. > > I managed to flash the "stable" firmware via arcconf, so I now have > 15753 ... > > I submitted a ticket there. They only offer support for money. I think I tried enough. Waiting for the smaller drive now. Thanks for your help, all.
Re: [gentoo-user] RAID: new drive on aac raid
Am 06.10.20 um 15:08 schrieb k...@aspodata.se: > Stefan G. Weichinger: >> I know the model: ICP5165BR > > https://ask.adaptec.com/app/answers/detail/a_id/17414/~/support-for-sata-and-sas-disk-drives-with-a-size-of-2tb-or-greater > > says it is supported up to 8TB drives using firmware v5.2.0 Build 17343 ** > > ** Firmware v5.2.0 Build 17343 for the ICP5045BL, ICP5085BL, ICP5805BL, >ICP5125BR, and ICP5165BR: Adaptec is providing minimally tested >firmware packages. Please contact Adaptec by PMC Technical Support >to obtain these firmware files. Have the TSID or serial number of >the product at hand when contacting support. Yes, I saw that as well. I managed to flash the "stable" firmware via arcconf, so I now have 15753 ... I submitted a ticket there. In the meantime we already ordered a smaller drive as well.
Re: [gentoo-user] RAID: new drive on aac raid
Stefan G. Weichinger: > Am 06.10.20 um 11:52 schrieb k...@aspodata.se: > > Stefan G. Weichinger: > >> Am 05.10.20 um 21:32 schrieb k...@aspodata.se: > > ... > >> What do you think, is 2 TB maybe too big for the controller? > > > 0a:0e.0 RAID bus controller: Adaptec AAC-RAID > > > > This doesn't really tells us which controller it is, try with > > > > lspci -s 0a:0e.0 -nn > > I know the model: ICP5165BR https://ask.adaptec.com/app/answers/detail/a_id/17414/~/support-for-sata-and-sas-disk-drives-with-a-size-of-2tb-or-greater says it is supported up to 8TB drives using firmware v5.2.0 Build 17343 ** ** Firmware v5.2.0 Build 17343 for the ICP5045BL, ICP5085BL, ICP5805BL, ICP5125BR, and ICP5165BR: Adaptec is providing minimally tested firmware packages. Please contact Adaptec by PMC Technical Support to obtain these firmware files. Have the TSID or serial number of the product at hand when contacting support. Regards, /Karl Hammar
Re: [gentoo-user] RAID: new drive on aac raid
On 05/10/2020 17:01, Stefan G. Weichinger wrote: Am 05.10.20 um 17:19 schrieb Stefan G. Weichinger: So my issue seems to be: non-working arcconf doesn't let me "enable" that one drive. Some kind of progress. Searched for more and older releases of arcconf, found Version 1.2 that doesn't crash here. This lets me view the physical device(s), but the new disk is marked as "Failed". Does it think the disk is a negative size? I looked, your Tosh is 2TB, and the other I looked at was 700GB. The raid website says a lot of older controllers can't cope with 2TB or larger disks ... Actually, the device information seems to confirm that - Total Size 0 MB ??? # ./arcconf GETCONFIG 1 PD | more Controllers found: 1 -- Physical Device information -- Device #0 Device is a Hard drive State : Failed Block Size : Unknown Supported : Yes Transfer Speed : Failed Reported Channel,Device(T:L) : 0,0(0:0) Reported Location : Connector 0, Device 0 Vendor : TOSHIBA Model : MG04SCA20EE Firmware : 0104 Serial number : 30A0A00UFX2B World-wide name: 539A08327484 Total Size : 0 MB Write Cache: Unknown FRU: None S.M.A.R.T. : No S.M.A.R.T. warnings: 0 SSD: No
Re: [gentoo-user] RAID: new drive on aac raid
Am 06.10.20 um 11:52 schrieb k...@aspodata.se: > Some guesses: > > https://wiki.debian.org/LinuxRaidForAdmins#aacraid > says that it requires libstd++5 > > arcconf might fork and exec, one could try with strace and try to > see what happens > > one could, if the old suse dist. is available in a subdir, to chroot > to that sudir, and try arcconf from there Hmm, yes. Currently I think it's the ancient firmware ... and maybe arcconf also crashes when it's not matching some minimum version of firmware. As mentioned in my other reply, I wait for that formatting to finish and then I want to try a firmware upgrade (to brand new 2008 ;-) ).
Re: [gentoo-user] RAID: new drive on aac raid
Am 06.10.20 um 11:52 schrieb k...@aspodata.se: > Stefan G. Weichinger: >> Am 05.10.20 um 21:32 schrieb k...@aspodata.se: > ... >> What do you think, is 2 TB maybe too big for the controller? > 0a:0e.0 RAID bus controller: Adaptec AAC-RAID > > This doesn't really tells us which controller it is, try with > > lspci -s 0a:0e.0 -nn I know the model: ICP5165BR with ancient Firmware. Currently I am in the Controller's BIOS or however you call that. I try to initialize and/or format the drive to make it available. The format runs for hours already. I prepared a FreeDOS iso with the latest firmware to flash ... after the formatting is done. Maybe then the compatibility is better. Or the controller becomes a paperweight. I have the impression that the controller hasn't yet fully recognized the disk, it is displayed differently in the array and disk menus of the firmware UI. >>> What do sg_verify /dev/sg11 return ? >> nothing > > Well, you have to check the return status: echo $? Maybe later, right now not possible (as mentioned above).
Re: [gentoo-user] RAID: new drive on aac raid
Stefan G. Weichinger: > Am 05.10.20 um 16:38 schrieb k...@aspodata.se: ... > But no luck with any version of arcconf so far. Unpacked several zips, > tried 2 releases, 32 and 64 bits .. all crash. > > > Just a poke in the dark, does ldd report all libs found, as in: > > $ ldd /bin/ls > > linux-vdso.so.1 (0x7ffcbab4c000) > > libc.so.6 => /lib64/libc.so.6 (0x7fece3ad5000) > > /lib64/ld-linux-x86-64.so.2 (0x7fece3d1c000) > > $ > > Yeah, that works. Some guesses: https://wiki.debian.org/LinuxRaidForAdmins#aacraid says that it requires libstd++5 arcconf might fork and exec, one could try with strace and try to see what happens one could, if the old suse dist. is available in a subdir, to chroot to that sudir, and try arcconf from there Regards, /Karl Hammar
Re: [gentoo-user] RAID: new drive on aac raid
Stefan G. Weichinger:
> Am 05.10.20 um 21:32 schrieb k...@aspodata.se:
...
> What do you think, is 2 TB maybe too big for the controller?
>>> 0a:0e.0 RAID bus controller: Adaptec AAC-RAID
This doesn't really tells us which controller it is, try with
lspci -s 0a:0e.0 -nn
In the kernel source one can then find drivers/scsi/aacraid/linit.h,
and since lsscsi says ICP, my guess it is one of theese below.
When we know which one, one can try to check for issues.
/*
* Because of the way Linux names scsi devices, the order in this table has
* become important. Check for on-board Raid first, add-in cards second.
*
* Note: The last field is used to index into aac_drivers below.
*/
static const struct pci_device_id aac_pci_tbl[] = {
...
{ 0x9005, 0x0286, 0x9005, 0x029e, 0, 0, 25 }, /* ICP9024RO (Lancer) */
{ 0x9005, 0x0286, 0x9005, 0x029f, 0, 0, 26 }, /* ICP9014RO (Lancer) */
{ 0x9005, 0x0286, 0x9005, 0x02a0, 0, 0, 27 }, /* ICP9047MA (Lancer) */
{ 0x9005, 0x0286, 0x9005, 0x02a1, 0, 0, 28 }, /* ICP9087MA (Lancer) */
{ 0x9005, 0x0286, 0x9005, 0x02a3, 0, 0, 29 }, /* ICP5445AU
(Hurricane44) */
{ 0x9005, 0x0285, 0x9005, 0x02a4, 0, 0, 30 }, /* ICP9085LI (Marauder-X)
*/
{ 0x9005, 0x0285, 0x9005, 0x02a5, 0, 0, 31 }, /* ICP5085BR (Marauder-E)
*/
{ 0x9005, 0x0286, 0x9005, 0x02a6, 0, 0, 32 }, /* ICP9067MA (Intruder-6)
*/
...
};
MODULE_DEVICE_TABLE(pci, aac_pci_tbl);
/*
* dmb - For now we add the number of channels to this structure.
* In the future we should add a fib that reports the number of channels
* for the card. At that time we can remove the channels from here
*/
static struct aac_driver_ident aac_drivers[] = {
...
{ aac_rkt_init, "aacraid", "ICP ", "ICP9024RO ", 2 }, /*
ICP9024RO (Lancer) */
{ aac_rkt_init, "aacraid", "ICP ", "ICP9014RO ", 1 }, /*
ICP9014RO (Lancer) */
{ aac_rkt_init, "aacraid", "ICP ", "ICP9047MA ", 1 }, /*
ICP9047MA (Lancer) */
{ aac_rkt_init, "aacraid", "ICP ", "ICP9087MA ", 1 }, /*
ICP9087MA (Lancer) */
{ aac_rkt_init, "aacraid", "ICP ", "ICP5445AU ", 1 }, /*
ICP5445AU (Hurricane44) */
{ aac_rx_init, "aacraid", "ICP ", "ICP9085LI ", 1 }, /*
ICP9085LI (Marauder-X) */
{ aac_rx_init, "aacraid", "ICP ", "ICP5085BR ", 1 }, /*
ICP5085BR (Marauder-E) */
{ aac_rkt_init, "aacraid", "ICP ", "ICP9067MA ", 1 }, /*
ICP9067MA (Intruder-6) */
...
};
> > What do sg_verify /dev/sg11 return ?
> nothing
Well, you have to check the return status: echo $?
> > Can you do sg_dd if=foo of=/dev/sg11 count=10 and get it back with
> > sg_dd if=/dev/sg11 of=bar count=10, with cmp foo bar; echo $?
> > returning 0 ?
> Yes, that works.
Then it seems the drive itself is ok.
Regards,
/Karl Hammar
Re: [gentoo-user] RAID: new drive on aac raid
Am 05.10.20 um 21:32 schrieb k...@aspodata.se: > What if you put it on the 53c1030 card, can you do that, at least to > verify the disk ? I am 600kms away from that server and the people I could send to the basement there aren't very competent in these things. I am afraid that won't work out well. I only told them to remove and re-insert the new drive. Maybe some contact issue. What do you think, is 2 TB maybe too big for the controller? > What do sg_verify /dev/sg11 return ? nothing > Can you do sg_dd if=foo of=/dev/sg11 count=10 and get it back with > sg_dd if=/dev/sg11 of=bar count=10, with cmp foo bar; echo $? > returning 0 ? Yes, that works.
Re: [gentoo-user] RAID: new drive on aac raid
Stefan G. Weichinger: ... > Searched for more and older releases of arcconf, found Version 1.2 that > doesn't crash here. > > This lets me view the physical device(s), but the new disk is marked as > "Failed". ... What if you put it on the 53c1030 card, can you do that, at least to verify the disk ? What do sg_verify /dev/sg11 return ? Can you do sg_dd if=foo of=/dev/sg11 count=10 and get it back with sg_dd if=/dev/sg11 of=bar count=10, with cmp foo bar; echo $? returning 0 ? Regards, /Karl Hammar
Re: [gentoo-user] RAID: new drive on aac raid
Am 05.10.20 um 17:19 schrieb Stefan G. Weichinger: > So my issue seems to be: non-working arcconf doesn't let me "enable" > that one drive. Some kind of progress. Searched for more and older releases of arcconf, found Version 1.2 that doesn't crash here. This lets me view the physical device(s), but the new disk is marked as "Failed". # ./arcconf GETCONFIG 1 PD | more Controllers found: 1 -- Physical Device information -- Device #0 Device is a Hard drive State : Failed Block Size : Unknown Supported : Yes Transfer Speed : Failed Reported Channel,Device(T:L) : 0,0(0:0) Reported Location : Connector 0, Device 0 Vendor : TOSHIBA Model : MG04SCA20EE Firmware : 0104 Serial number : 30A0A00UFX2B World-wide name: 539A08327484 Total Size : 0 MB Write Cache: Unknown FRU: None S.M.A.R.T. : No S.M.A.R.T. warnings: 0 SSD: No Tried a rescan and a clearing of status ... no change. Maybe the disk is too big for that controller. Creating a LD also fails: # ./arcconf CREATE 1 LOGICALDRIVE MAX volume 0,0 Controllers found: 1 A selected device is not available for use. Command aborted. juno /usr/portage/distfiles/linux_x64 # ./arcconf CREATE 1 LOGICALDRIVE MAX volume 0 0 Controllers found: 1 A selected device is not available for use. Command aborted. ... annoying and frustrating
Re: [gentoo-user] RAID: new drive on aac raid
Am 05.10.20 um 16:57 schrieb Rich Freeman: > If you're doing software RAID or just individual disks, then you're > probably going to go into the controller and basically configure that > disk as standalone, or as a 1-disk "RAID". That will make it appear > to the OS, and then you can do whatever you want with it at the OS > level (stick a filesystem on it, put it in a RAID/lvm, whatever). > > I find this sort of thing really annoying. Same here! ;-) > I prefer HBAs that just do > IT mode or equivalent - acting as a dumb HBA and passing all the > drives through to the OS. It isn't that it doesn't work - it is just > that you're now married to that HBA card vendor and if anything > happens to the card you have to replace it with something compatible > and reconfigure it using their software/etc, or else all your data is > unreadable. Even if you have backups it isn't something you want to > just have to deal with if you're talking about a lot of data. Yep. So my issue seems to be: non-working arcconf doesn't let me "enable" that one drive. I *might* consider booting up the older Suse OS (still somewhere as well) via the flaky old Java-KVM and try things there. The server is ~600kms away, so my possibilities with Live-USB-sticks etc are limited right now.
Re: [gentoo-user] RAID: new drive on aac raid
Am 05.10.20 um 16:38 schrieb k...@aspodata.se: > And theese on the aac, since they have the same scsi host, and I guess > that scsi ch.0 is for the configured drives and ch.1 for the raw drives: >> [1:0:1:0]diskICP SAS2 V1.0 /dev/sda >> [1:0:2:0]diskICP Device 2 V1.0 /dev/sdb >> [1:0:3:0]diskICP Device 3 V1.0 /dev/sdc >> [1:0:4:0]diskICP Device 4 V1.0 /dev/sdd >> [1:0:5:0]diskICP Device 5 V1.0 /dev/sde >> [1:0:6:0]diskICP Device 6 V1.0 /dev/sdf >> [1:0:7:0]diskICP Device 7 V1.0 /dev/sdg >> [1:0:8:0]diskICP Device 8 V1.0 /dev/sdh >> [1:0:9:0]diskICP Device 9 V1.0 /dev/sdi >> [1:1:0:0]diskTOSHIBA MG04SCA20EE 0104 Thanks for your analysis and pointers! > Perhaps theese links will help: > > https://www.cyberciti.biz/faq/linux-checking-sas-sata-disks-behind-adaptec-raid-controllers/ > http://updates.aslab.com/doc/disk-controller/aacraid_guide.pdf > https://hwraid.le-vert.net/wiki/Adaptec Somehow. I get smartctl output for that disk: # smartctl -d scsi --all /dev/sg11 smartctl 7.0 2018-12-30 r4883 [x86_64-linux-4.14.83-gentoo-smp] (local build) Copyright (C) 2002-18, Bruce Allen, Christian Franke, www.smartmontools.org === START OF INFORMATION SECTION === Vendor: TOSHIBA Product: MG04SCA20EE Revision: 0104 Compliance: SPC-4 User Capacity:2.000.398.934.016 bytes [2,00 TB] Logical block size: 512 bytes Physical block size: 4096 bytes Rotation Rate:7200 rpm Form Factor: 3.5 inches Logical Unit id: 0x539a08327485 Serial number:30A0A00UFX2B Device type: disk Transport protocol: SAS (SPL-3) Local Time is:Mon Oct 5 18:54:44 2020 CEST SMART support is: Available - device has SMART capability. SMART support is: Disabled Temperature Warning: Disabled or Not Supported But no luck with any version of arcconf so far. Unpacked several zips, tried 2 releases, 32 and 64 bits .. all crash. > Just a poke in the dark, does ldd report all libs found, as in: > $ ldd /bin/ls > linux-vdso.so.1 (0x7ffcbab4c000) > libc.so.6 => /lib64/libc.so.6 (0x7fece3ad5000) > /lib64/ld-linux-x86-64.so.2 (0x7fece3d1c000) > $ Yeah, that works.
Re: [gentoo-user] RAID: new drive on aac raid
On Mon, Oct 5, 2020 at 10:38 AM wrote: > > Stefan G. Weichinger: > > On an older server the customer replaced a SAS drive. > > > > I see it as /dev/sg11, but not yes as /dev/sdX, it is not visible in "lsblk" > > Perhaps theese links will help: > > https://www.cyberciti.biz/faq/linux-checking-sas-sata-disks-behind-adaptec-raid-controllers/ > http://updates.aslab.com/doc/disk-controller/aacraid_guide.pdf > https://hwraid.le-vert.net/wiki/Adaptec > I don't know the details of any of these controllers, but you have the gist of it. The RAID controller is abstracting the individual drives and so the OS doesn't see them. You need to do at least some of the configuration through the controller. That usually requires vendor-specific software, which is often available for linux, and which in some cases is packaged for Gentoo. There are a lot of ways to do something like this. If you're doing hardware RAID you'd just replace/etc the disk in the raid (I'm actually surprised in this case that just swapping the drive in the same slot didn't already do this), and the hardware RAID will rebuild it, and the OS doesn't see anything at all. You might need the utility, but that is about it. If you're doing software RAID or just individual disks, then you're probably going to go into the controller and basically configure that disk as standalone, or as a 1-disk "RAID". That will make it appear to the OS, and then you can do whatever you want with it at the OS level (stick a filesystem on it, put it in a RAID/lvm, whatever). I find this sort of thing really annoying. I prefer HBAs that just do IT mode or equivalent - acting as a dumb HBA and passing all the drives through to the OS. It isn't that it doesn't work - it is just that you're now married to that HBA card vendor and if anything happens to the card you have to replace it with something compatible and reconfigure it using their software/etc, or else all your data is unreadable. Even if you have backups it isn't something you want to just have to deal with if you're talking about a lot of data. -- Rich
Re: [gentoo-user] RAID: new drive on aac raid
Stefan G. Weichinger: > On an older server the customer replaced a SAS drive. > > I see it as /dev/sg11, but not yes as /dev/sdX, it is not visible in "lsblk" ... Not that I think it will help you much, but there is sys-apps/sg3_utils: # lsscsi [0:0:0:0]diskATA TOSHIBA MG03ACA3 FL1A /dev/sda [1:0:0:0]diskATA TOSHIBA MG03ACA3 FL1A /dev/sdb [5:0:0:0]cd/dvd ASUS DRW-24F1ST a 1.00 /dev/scd0 # sg_map -x -i /dev/sg0 0 0 0 0 0 /dev/sda ATA TOSHIBA MG03ACA3 FL1A /dev/sg1 1 0 0 0 0 /dev/sdb ATA TOSHIBA MG03ACA3 FL1A /dev/sg2 5 0 0 0 5 /dev/scd0 ASUS DRW-24F1ST a1.00 # sg_inq /dev/sg1 standard INQUIRY: PQual=0 Device_type=0 RMB=0 LU_CONG=0 version=0x05 [SPC-3] [AERC=0] [TrmTsk=0] NormACA=0 HiSUP=0 Resp_data_format=2 SCCS=0 ACC=0 TPGS=0 3PC=0 Protect=0 [BQue=0] EncServ=0 MultiP=0 [MChngr=0] [ACKREQQ=0] Addr16=0 [RelAdr=0] WBus16=0 Sync=0 [Linked=0] [TranDis=0] CmdQue=1 [SPI: Clocking=0x0 QAS=0 IUS=0] length=96 (0x60) Peripheral device type: disk Vendor identification: ATA Product identification: TOSHIBA MG03ACA3 Product revision level: FL1A Unit serial number:14UAKBPDF You have theese two: > 07:01.0 SCSI storage controller: LSI Logic / Symbios Logic 53c1030 PCI-X > 0a:0e.0 RAID bus controller: Adaptec AAC-RAID I guess that the tape drive and mediachanger is on the 53c1030: > [0:0:1:0]tapeHP Ultrium 4-SCSI B12H /dev/st0 > [0:0:1:1]mediumx OVERLAND NEO Series 0510 /dev/sch0 And theese on the aac, since they have the same scsi host, and I guess that scsi ch.0 is for the configured drives and ch.1 for the raw drives: > [1:0:1:0]diskICP SAS2 V1.0 /dev/sda > [1:0:2:0]diskICP Device 2 V1.0 /dev/sdb > [1:0:3:0]diskICP Device 3 V1.0 /dev/sdc > [1:0:4:0]diskICP Device 4 V1.0 /dev/sdd > [1:0:5:0]diskICP Device 5 V1.0 /dev/sde > [1:0:6:0]diskICP Device 6 V1.0 /dev/sdf > [1:0:7:0]diskICP Device 7 V1.0 /dev/sdg > [1:0:8:0]diskICP Device 8 V1.0 /dev/sdh > [1:0:9:0]diskICP Device 9 V1.0 /dev/sdi > [1:1:0:0]diskTOSHIBA MG04SCA20EE 0104 > [1:1:1:0]diskSEAGATE ST373455SS 0002 - > [1:1:3:0]diskWDC WD7500AZEX-00RKK 0A80 - > [1:1:4:0]diskWDC WD7500AZEX-00RKK 0A80 - > [1:1:5:0]diskWDC WD7500AZEX-00RKK 0A80 - > [1:1:6:0]diskWDC WD7500AZEX-00BN5 1A01 - > [1:1:7:0]diskWDC WD7500AZEX-00BN5 1A01 - > [1:1:8:0]diskWDC WD7500AZEX-00RKK 0A80 - > [1:1:9:0]diskST375052 8AS CC44 - > [1:1:10:0] diskWDC WD7500AZEX-00BN5 1A01 - > [1:1:11:0] diskWDC WD7500AZEX-00BN5 1A01 - Perhaps theese links will help: https://www.cyberciti.biz/faq/linux-checking-sas-sata-disks-behind-adaptec-raid-controllers/ http://updates.aslab.com/doc/disk-controller/aacraid_guide.pdf https://hwraid.le-vert.net/wiki/Adaptec Just a poke in the dark, does ldd report all libs found, as in: $ ldd /bin/ls linux-vdso.so.1 (0x7ffcbab4c000) libc.so.6 => /lib64/libc.so.6 (0x7fece3ad5000) /lib64/ld-linux-x86-64.so.2 (0x7fece3d1c000) $ Regards, /Karl Hammar
