Dmitry Morozovsky
2012-Mar-18 16:11 UTC
ahci hangs on Supermicro MicroCloud second channel
Dear colleagues, I've start testing SuperMicro MicroCloud[1] to have high-density routers cluster, and experiencing strange effects with disk subsystem: - on stable/8, it does detect AHCI controller, but detects disks as non-ahci ad* - on stable/9, disks are shown as ada*, but disk on second channel has constant read/write hangs, showing 100% load on few hundreds kBps in gstat. disk controller is Intel C204 PCH: ahci0: <Intel Cougar Point AHCI SATA controller> port 0xf050-0xf057,0xf040-0xf043,0xf030-0xf037,0xf020-0xf023,0xf000-0xf01f mem 0xfa901000-0xfa9017ff irq 19 at device 31.2 on pci0 ahci0: attempting to allocate 1 MSI vectors (1 supported) msi: routing MSI IRQ 284 to local APIC 0 vector 81 ahci0: using IRQ 284 for MSI ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier not supported ahci0: Caps: 64bit NCQ SNTF ALP AL CLO 6Gbps PMD SSC PSC 32cmd EM 6ports ahci0: Caps2: APST ahci0: EM Caps: ALHD XMT SMB LED ahcich0: <AHCI channel> at channel 0 on ahci0 ahcich0: Caps: ahcich1: <AHCI channel> at channel 1 on ahci0 ahcich1: Caps: pass0 at ahcich0 bus 0 scbus0 target 0 lun 0 pass0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device pass0: Serial Number WD-WCAYUFH26175 pass0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) pass0: Command Queueing enabled pass1 at ahcich1 bus 0 scbus1 target 0 lun 0 pass1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device pass1: Serial Number WD-WCAYUFH32290 pass1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) pass1: Command Queueing enabled ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 ada0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device ada0: Serial Number WD-WCAYUFH26175 ada0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) ada0: Command Queueing enabled ada0: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) ada0: Previously was known as ad4 ada1 at ahcich1 bus 0 scbus1 target 0 lun 0 GEOM: new disk ada0 GEOM: new disk ada1 ada1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device ada1: Serial Number WD-WCAYUFH32290 ada1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) ada1: Command Queueing enabled ada1: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) ada1: Previously was known as ad6 Any hints? [1] http://www.supermicro.nl/products/system/3U/5037/SYS-5037MC-H8TRF.cfm -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck@FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------
Hi, Did you check whether there's newer firmware for the microcloud mainboards? Does the integrated ctrls are in AHCI mode in the BIOS? You may also ask Supermicro if it turns out that it's not a FreeBSD problem, but be prepared that they'll ask for enterprise drives first. Regards, Andras On Sun, 18 Mar 2012 20:10:34 +0400 (MSK), Dmitry Morozovsky wrote:> Dear colleagues, > > I've start testing SuperMicro MicroCloud[1] to have high-density > routers > cluster, and experiencing strange effects with disk subsystem: > > - on stable/8, it does detect AHCI controller, but detects disks as > non-ahci > ad* > - on stable/9, disks are shown as ada*, but disk on second channel > has constant > read/write hangs, showing 100% load on few hundreds kBps in gstat. > > disk controller is Intel C204 PCH: > > ahci0: <Intel Cougar Point AHCI SATA controller> port > 0xf050-0xf057,0xf040-0xf043,0xf030-0xf037,0xf020-0xf023,0xf000-0xf01f > mem > 0xfa901000-0xfa9017ff irq 19 at device 31.2 on pci0 > ahci0: attempting to allocate 1 MSI vectors (1 supported) > msi: routing MSI IRQ 284 to local APIC 0 vector 81 > ahci0: using IRQ 284 for MSI > ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier not supported > ahci0: Caps: 64bit NCQ SNTF ALP AL CLO 6Gbps PMD SSC PSC 32cmd EM > 6ports > ahci0: Caps2: APST > ahci0: EM Caps: ALHD XMT SMB LED > ahcich0: <AHCI channel> at channel 0 on ahci0 > ahcich0: Caps: > ahcich1: <AHCI channel> at channel 1 on ahci0 > ahcich1: Caps: > > pass0 at ahcich0 bus 0 scbus0 target 0 lun 0 > pass0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > pass0: Serial Number WD-WCAYUFH26175 > pass0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > pass0: Command Queueing enabled > pass1 at ahcich1 bus 0 scbus1 target 0 lun 0 > pass1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > pass1: Serial Number WD-WCAYUFH32290 > pass1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > pass1: Command Queueing enabled > ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 > ada0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > ada0: Serial Number WD-WCAYUFH26175 > ada0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada0: Command Queueing enabled > ada0: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) > ada0: Previously was known as ad4 > ada1 at ahcich1 bus 0 scbus1 target 0 lun 0 > GEOM: new disk ada0 > GEOM: new disk ada1 > ada1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > ada1: Serial Number WD-WCAYUFH32290 > ada1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada1: Command Queueing enabled > ada1: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) > ada1: Previously was known as ad6 > > Any hints? > > > [1] > http://www.supermicro.nl/products/system/3U/5037/SYS-5037MC-H8TRF.cfm
Hi, And another question, have you tried with other type and vendor of disks? We saw some strange behavior with specific disks hanged onto a specific ctrl, but that was with Linux. :) Andras On Sun, 18 Mar 2012 20:10:34 +0400 (MSK), Dmitry Morozovsky wrote:> Dear colleagues, > > I've start testing SuperMicro MicroCloud[1] to have high-density > routers > cluster, and experiencing strange effects with disk subsystem: > > - on stable/8, it does detect AHCI controller, but detects disks as > non-ahci > ad* > - on stable/9, disks are shown as ada*, but disk on second channel > has constant > read/write hangs, showing 100% load on few hundreds kBps in gstat. > > disk controller is Intel C204 PCH: > > ahci0: <Intel Cougar Point AHCI SATA controller> port > 0xf050-0xf057,0xf040-0xf043,0xf030-0xf037,0xf020-0xf023,0xf000-0xf01f > mem > 0xfa901000-0xfa9017ff irq 19 at device 31.2 on pci0 > ahci0: attempting to allocate 1 MSI vectors (1 supported) > msi: routing MSI IRQ 284 to local APIC 0 vector 81 > ahci0: using IRQ 284 for MSI > ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier not supported > ahci0: Caps: 64bit NCQ SNTF ALP AL CLO 6Gbps PMD SSC PSC 32cmd EM > 6ports > ahci0: Caps2: APST > ahci0: EM Caps: ALHD XMT SMB LED > ahcich0: <AHCI channel> at channel 0 on ahci0 > ahcich0: Caps: > ahcich1: <AHCI channel> at channel 1 on ahci0 > ahcich1: Caps: > > pass0 at ahcich0 bus 0 scbus0 target 0 lun 0 > pass0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > pass0: Serial Number WD-WCAYUFH26175 > pass0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > pass0: Command Queueing enabled > pass1 at ahcich1 bus 0 scbus1 target 0 lun 0 > pass1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > pass1: Serial Number WD-WCAYUFH32290 > pass1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > pass1: Command Queueing enabled > ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 > ada0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > ada0: Serial Number WD-WCAYUFH26175 > ada0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada0: Command Queueing enabled > ada0: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) > ada0: Previously was known as ad4 > ada1 at ahcich1 bus 0 scbus1 target 0 lun 0 > GEOM: new disk ada0 > GEOM: new disk ada1 > ada1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > ada1: Serial Number WD-WCAYUFH32290 > ada1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada1: Command Queueing enabled > ada1: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) > ada1: Previously was known as ad6 > > Any hints? > > > [1] > http://www.supermicro.nl/products/system/3U/5037/SYS-5037MC-H8TRF.cfm
----- Original Message ----- From: "Dmitry Morozovsky" <marck@rinet.ru> To: <freebsd-stable@FreeBSD.org> Cc: <mav@FreeBSD.org> Sent: Sunday, March 18, 2012 4:10 PM Subject: ahci hangs on Supermicro MicroCloud second channel> Dear colleagues, > > I've start testing SuperMicro MicroCloud[1] to have high-density routers > cluster, and experiencing strange effects with disk subsystem: > > - on stable/8, it does detect AHCI controller, but detects disks as non-ahci > ad* > - on stable/9, disks are shown as ada*, but disk on second channel has constant > read/write hangs, showing 100% load on few hundreds kBps in gstat. > > disk controller is Intel C204 PCH: > > ahci0: <Intel Cougar Point AHCI SATA controller> port > 0xf050-0xf057,0xf040-0xf043,0xf030-0xf037,0xf020-0xf023,0xf000-0xf01f mem > 0xfa901000-0xfa9017ff irq 19 at device 31.2 on pci0 > ahci0: attempting to allocate 1 MSI vectors (1 supported) > msi: routing MSI IRQ 284 to local APIC 0 vector 81 > ahci0: using IRQ 284 for MSI > ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier not supported > ahci0: Caps: 64bit NCQ SNTF ALP AL CLO 6Gbps PMD SSC PSC 32cmd EM 6ports > ahci0: Caps2: APST > ahci0: EM Caps: ALHD XMT SMB LED > ahcich0: <AHCI channel> at channel 0 on ahci0 > ahcich0: Caps: > ahcich1: <AHCI channel> at channel 1 on ahci0 > ahcich1: Caps: > > pass0 at ahcich0 bus 0 scbus0 target 0 lun 0 > pass0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > pass0: Serial Number WD-WCAYUFH26175 > pass0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > pass0: Command Queueing enabled > pass1 at ahcich1 bus 0 scbus1 target 0 lun 0 > pass1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > pass1: Serial Number WD-WCAYUFH32290 > pass1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > pass1: Command Queueing enabled > ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 > ada0: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > ada0: Serial Number WD-WCAYUFH26175 > ada0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada0: Command Queueing enabled > ada0: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) > ada0: Previously was known as ad4 > ada1 at ahcich1 bus 0 scbus1 target 0 lun 0 > GEOM: new disk ada0 > GEOM: new disk ada1 > ada1: <WDC WD5000AAKX-001CA0 15.01H15> ATA-8 SATA 3.x device > ada1: Serial Number WD-WCAYUFH32290 > ada1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada1: Command Queueing enabled > ada1: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) > ada1: Previously was known as ad6 > > Any hints? > > > [1] http://www.supermicro.nl/products/system/3U/5037/SYS-5037MC-H8TRF.cfmWe have quite a few of these running 8.2-RELEASE-p6 on AHCI with no problems (kernel compiled with:- device ahci) ahci0: <Intel Cougar Point AHCI SATA controller> port 0xf050-0xf057,0xf040-0xf043,0xf030-0xf037,0xf020-0xf023,0xf000-0xf01f mem 0xfbc01000-0xfbc017ff irq 19 at device 31.2 on pci0 ahci0: [ITHREAD] ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier not supported ahcich0: <AHCI channel> at channel 0 on ahci0 ahcich0: [ITHREAD] ahcich1: <AHCI channel> at channel 1 on ahci0 ahcich1: [ITHREAD] ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 ada0: <WDC WD1003FBYX-01Y7B0 01.01V01> ATA-8 SATA 2.x device ada0: 300.000MB/s transfers (SATA 2.x, UDMA6, PIO 8192bytes) ada0: Command Queueing enabled ada0: 953869MB (1953525168 512 byte sectors: 16H 63S/T 16383C) ada1 at ahcich1 bus 0 scbus1 target 0 lun 0 ada1: <Corsair Force 3 SSD 1.3> ATA-8 SATA 3.x device ada1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) ada1: Command Queueing enabled ada1: 57241MB (117231408 512 byte sectors: 16H 63S/T 16383C) Given this might be worth seeing if 8.2 on AHCI fixes the hangs and hence its a regression in 9. If your running generic you should be able to just add the following to /boot/loader.conf ahci_load="YES" Regards Steve ===============================================This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmaster@multiplay.co.uk.
Dmitry Morozovsky
2012-Mar-18 17:08 UTC
ahci hangs on Supermicro MicroCloud second channel
Steven, On Sun, 18 Mar 2012, Steven Hartland wrote: [snip]> We have quite a few of these running 8.2-RELEASE-p6 on AHCI with no > problems (kernel compiled with:- device ahci) > > ahci0: <Intel Cougar Point AHCI SATA controller> port > 0xf050-0xf057,0xf040-0xf043,0xf030-0xf037,0xf020-0xf023,0xf000-0xf01f mem > 0xfbc01000-0xfbc017ff irq 19 at device 31.2 on pci0 > ahci0: [ITHREAD] > ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier not supported > ahcich0: <AHCI channel> at channel 0 on ahci0 > ahcich0: [ITHREAD] > ahcich1: <AHCI channel> at channel 1 on ahci0 > ahcich1: [ITHREAD] > > ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 > ada0: <WDC WD1003FBYX-01Y7B0 01.01V01> ATA-8 SATA 2.x device > ada0: 300.000MB/s transfers (SATA 2.x, UDMA6, PIO 8192bytes) > ada0: Command Queueing enabled > ada0: 953869MB (1953525168 512 byte sectors: 16H 63S/T 16383C) > ada1 at ahcich1 bus 0 scbus1 target 0 lun 0 > ada1: <Corsair Force 3 SSD 1.3> ATA-8 SATA 3.x device > ada1: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 8192bytes) > ada1: Command Queueing enabled > ada1: 57241MB (117231408 512 byte sectors: 16H 63S/T 16383C) > > Given this might be worth seeing if 8.2 on AHCI fixes the > hangs and hence its a regression in 9. If your running > generic you should be able to just add the following to > /boot/loader.conf > ahci_load="YES"Yes, I did it, but all I have ... wait a bit... Yes, I missed ahci.ko module on my PXE server in amd64/8 tree :-/ [kernel reinstall] Well, ahci problem solved, but I still have much worse performance (and different on ada0 and ada1!): ada0, MC 50-60 MBps ada1, MC 13-25 MBps ada*, 5017 130+ MBps Could you please post SATA/AHCI BIOS settings from your machines? Thanks a lot! -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck@FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------
----- Original Message ----- From: "Dmitry Morozovsky" <marck@rinet.ru>> Well, ahci problem solved, but I still have much worse performance (and > different on ada0 and ada1!): > ada0, MC 50-60 MBps > ada1, MC 13-25 MBps > ada*, 5017 130+ MBps > > Could you please post SATA/AHCI BIOS settings from your machines? Thanks a lot!Just AHCI configured as far as I remember. What you mean by MC vs 5017? How are you measuring disk speed? What value do you have for sysctl vfs.read_max? Regards Steve ===============================================This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmaster@multiplay.co.uk.
Dmitry Morozovsky
2012-Mar-18 20:45 UTC
ahci hangs on Supermicro MicroCloud second channel
On Sun, 18 Mar 2012, Steven Hartland wrote:> > different on ada0 and ada1!): > > ada0, MC 50-60 MBps ada1, MC 13-25 MBps > > ada*, 5017 130+ MBps > > > > Could you please post SATA/AHCI BIOS settings from your machines? Thanks a > > lot! > > Just AHCI configured as far as I remember.Microcloud has AFAIR additional settings for delaying disk spin and enablink/disabling hotswap, but mangling these options does not change anything for me> What you mean by MC vs 5017?platform name: MicroCloud vs 5017C-MTF, both based on the same chipset> How are you measuring disk speed?the simplest: dd if=/dev/ada0 of=/dev/null bs=1m count=16k (linear read 16g at the beginning of disk)> What value do you have for sysctl vfs.read_max?default for both cases, 8 -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck@FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------