Re: IO fluctuation
On Saturday 21 August 2010 19:48:32 Steffen Neubauer wrote: > I think I can exclude the ZFS implementation too, because I tried dd > if=/dev/da0 of=/dev/null and executed killall -SIGINFO dd while it was > copying and it looked like it stalled randomly too. Not related your problem, but you can press ^T to send SIGINFO to the currently running process. Saves some typing :-) - Pieter ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: IO fluctuation
http://forums.freebsd.org/showthread.php?t=17036&; This is a good starting point, there is also some good info using certain tools to force the drives to not goto spindown sleep. On Sun, Aug 22, 2010 at 4:29 AM, Steffen Neubauer wrote: > Hello Claudiu :) > > Sorry that I didn't answer these obvious questions in my first post. > > Am 21.08.2010 20:16, schrieb claudiu vasadi: > > How did you setup your drives in ZFS ? stripe ? >> > > raidz1 Here is my zpool status output: > > # zpool status > pool: tank > state: ONLINE > scrub: none requested > config: > >NAMESTATE READ WRITE > CKSUM >tankONLINE 0 0 > 0 > raidz1ONLINE 0 0 > 0 >gptid/7b91a9b4-ab05-11df-8c6e-001d7d004144 ONLINE 0 0 > 0 >gptid/7cebaf7b-ab05-11df-8c6e-001d7d004144 ONLINE 0 0 > 0 >gptid/7e43d7a2-ab05-11df-8c6e-001d7d004144 ONLINE 0 0 > 0 > > errors: No known data errors > > > what bsd version do you have and what platform ? >> > > # uname -a > FreeBSD tank 8.1-RELEASE FreeBSD 8.1-RELEASE #0: Mon Jul 19 02:36:49 UTC > 2010 r...@mason.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC amd64 > > > > Do you see this big fluctuation when copying some big file like say and >> iso >> or movie ? >> > > Both, on small filesystem access and when copying big files. > When I'm listening to music with VLC over NFS it stalls regularly. > When I'm working on the shell the commands block randomly for a certain > time. > > But I get that fluctuation on raw device access, too, look: > > # dd if=/dev/ad0 of=/dev/zero bs=1M > ^C0+0 records in > 0+0 records out > 0 bytes transferred in 9.848682 secs (0 bytes/sec) > # dd if=/dev/ad0 of=/dev/zero bs=1M > ^C1056+0 records in > 1056+0 records out > 1107296256 bytes transferred in 9.492375 secs (116651130 bytes/sec) > > So this is not a zfs problem. But I don't think it's the hardware, too - I > used that mainboard before FreeBSD with no problems. > And the WD EARS drives can't be the problem on raw device access, too, > because I used bs=1M, that should be enough... > > This is really strange... > > Greetings, > Steffen > > ___ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to " > freebsd-questions-unsubscr...@freebsd.org" > -- Regards, Jason Leschnik. Mob. 0432 35 4224 Uni mail. jml...@uow.edu.au Work mail. j...@ansto.gov.au ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: IO fluctuation
On Sat, 21 Aug 2010 20:29:05 +0200 Steffen Neubauer wrote: > So this is not a zfs problem. But I don't think it's the hardware, > too - I used that mainboard before FreeBSD with no problems. > And the WD EARS drives can't be the problem on raw device access, > too, because I used bs=1M, that should be enough... It _is_ the EARS: the hang is probably the drive spinning back up after it's turned itself off - it has really aggressive power management by default. For more information about the variation, see http://maycontaintracesofbolts.blogspot.com/2010/08/benchmarking-advanced-format-drives.html -- Bruce Cran ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: IO fluctuation
Hello Claudiu :) Sorry that I didn't answer these obvious questions in my first post. Am 21.08.2010 20:16, schrieb claudiu vasadi: How did you setup your drives in ZFS ? stripe ? raidz1 Here is my zpool status output: # zpool status pool: tank state: ONLINE scrub: none requested config: NAMESTATE READ WRITE CKSUM tankONLINE 0 0 0 raidz1ONLINE 0 0 0 gptid/7b91a9b4-ab05-11df-8c6e-001d7d004144 ONLINE 0 0 0 gptid/7cebaf7b-ab05-11df-8c6e-001d7d004144 ONLINE 0 0 0 gptid/7e43d7a2-ab05-11df-8c6e-001d7d004144 ONLINE 0 0 0 errors: No known data errors what bsd version do you have and what platform ? # uname -a FreeBSD tank 8.1-RELEASE FreeBSD 8.1-RELEASE #0: Mon Jul 19 02:36:49 UTC 2010 r...@mason.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC amd64 Do you see this big fluctuation when copying some big file like say and iso or movie ? Both, on small filesystem access and when copying big files. When I'm listening to music with VLC over NFS it stalls regularly. When I'm working on the shell the commands block randomly for a certain time. But I get that fluctuation on raw device access, too, look: # dd if=/dev/ad0 of=/dev/zero bs=1M ^C0+0 records in 0+0 records out 0 bytes transferred in 9.848682 secs (0 bytes/sec) # dd if=/dev/ad0 of=/dev/zero bs=1M ^C1056+0 records in 1056+0 records out 1107296256 bytes transferred in 9.492375 secs (116651130 bytes/sec) So this is not a zfs problem. But I don't think it's the hardware, too - I used that mainboard before FreeBSD with no problems. And the WD EARS drives can't be the problem on raw device access, too, because I used bs=1M, that should be enough... This is really strange... Greetings, Steffen ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: IO fluctuation
Hello Steffen, How did you setup your drives in ZFS ? stripe ? what bsd version do you have and what platform ? Do you see this big fluctuation when copying some big file like say and iso or movie ? ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
IO fluctuation
Hello, in the last few days I set up a freebsd box with 3 Western Digital Caviar Green 2TB drives (WD EARS series, but that shouldn't be the problem...) with ZFS on the drives. I'm experiencing really annoying fluctuations with the performance of the drives. Here is an example: http://www.stefreak.de/io-einbrueche.png (Downloading data through my ADSL) The problem occurs both locally (executing commands on the shell hangs randomly) and over the network, of course. So the problem can't be the networking stack or chip. I think I can exclude the ZFS implementation too, because I tried dd if=/dev/da0 of=/dev/null and executed killall -SIGINFO dd while it was copying and it looked like it stalled randomly too. Now I'm searching for the cause of the problem. Has anyone an idea? Here's my pciconf -lv output: # pciconf -lv hos...@pci0:0:0:0: class=0x06 card=0x50001458 chip=0x29c08086 rev=0x02 hdr=0x00 vendor = 'Intel Corporation' device = 'P35/G33 (Bearlake) Processor to I/O Controller' class = bridge subclass = HOST-PCI no...@pci0:0:27:0: class=0x040300 card=0xa0021458 chip=0x27d88086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = 'IDT High Definition Audio Driver (BA101897)' class = multimedia subclass = HDA pc...@pci0:0:28:0: class=0x060400 card=0x50011458 chip=0x27d08086 rev=0x01 hdr=0x01 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) PCIe Root Port' class = bridge subclass = PCI-PCI pc...@pci0:0:28:1: class=0x060400 card=0x50011458 chip=0x27d28086 rev=0x01 hdr=0x01 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) PCIe Root Port' class = bridge subclass = PCI-PCI uh...@pci0:0:29:0: class=0x0c0300 card=0x50041458 chip=0x27c88086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) USB Universal Host Controller' class = serial bus subclass = USB uh...@pci0:0:29:1: class=0x0c0300 card=0x50041458 chip=0x27c98086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) USB Universal Host Controller' class = serial bus subclass = USB uh...@pci0:0:29:2: class=0x0c0300 card=0x50041458 chip=0x27ca8086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) USB Universal Host Controller' class = serial bus subclass = USB uh...@pci0:0:29:3: class=0x0c0300 card=0x50041458 chip=0x27cb8086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) USB Universal Host Controller' class = serial bus subclass = USB eh...@pci0:0:29:7: class=0x0c0320 card=0x50061458 chip=0x27cc8086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801G (ICH7 Family) USB 2.0 Enhanced Host Controller' class = serial bus subclass = USB pc...@pci0:0:30:0: class=0x060401 card=0x50001458 chip=0x244e8086 rev=0xe1 hdr=0x01 vendor = 'Intel Corporation' device = '82801 Family (ICH2/3/4/5/6/7/8/9,63xxESB) Hub Interface to PCI Bridge' class = bridge subclass = PCI-PCI is...@pci0:0:31:0: class=0x060100 card=0x50011458 chip=0x27b88086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = 'Intel 82801GB/GR (ICH7 Family) LPC Interface Controller - 27B8 (945GL)' class = bridge subclass = PCI-ISA atap...@pci0:0:31:2:class=0x01018a card=0xb0021458 chip=0x27c08086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801GB/GR/GH (ICH7 Family) Serial ATA Storage Controller' class = mass storage subclass = ATA no...@pci0:0:31:3: class=0x0c0500 card=0x50011458 chip=0x27da8086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = 'Intel[R] 82801G (ICH7 Family) C- 27DA (82801G)' class = serial bus subclass = SMBus r...@pci0:2:0:0: class=0x02 card=0xe0001458 chip=0x816810ec rev=0x01 hdr=0x00 vendor = 'Realtek Semiconductor' device = 'Gigabit Ethernet NIC(NDIS 6.0) (RTL8168/8111/8111c)' class = network subclass = ethernet Greetings & Thank You, Steffen ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"