Re: mod_fcgid doesn't work in 9-stable jails after upgrade from 8.x
I could finally take the time to look into this, so here's the solution: Setting jail_sysvipc_allow="YES" in rc.conf is no longer enough, it only sets the sysctl, but the new rc.d/jail script won't add the allow.sysvipc=1 parameter. So in order to work, you must change the above to: jail_jailname_parameters="allow.sysvipc=1" ps: it's not related to 9, stable/8 rc.d/jail has the same new style jail invocation and hence the same problem. On 11/04/2012 07:47 PM, Attila Nagy wrote: Hi, I've just tried to upgrade a machine running an older 8-stable to 9-stable@r242549M without success. It runs an apache with mod_fcgid in a jail and the latter can't start with the error message of: [Sun Nov 04 16:09:12 2012] [emerg] (78)Function not implemented: mod_fcgid: Can't create shared memory for size 1192488 bytes security.jail.sysvipc_allowed is enabled (it was needed on 8.x too), nothing else has changed. There are some reports from this, but from earlier versions, and the only confirmed solution was sysvipc_allowed, which works for 8.x, but doesn't with the above version. Any ideas? ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
mod_fcgid doesn't work in 9-stable jails after upgrade from 8.x
Hi, I've just tried to upgrade a machine running an older 8-stable to 9-stable@r242549M without success. It runs an apache with mod_fcgid in a jail and the latter can't start with the error message of: [Sun Nov 04 16:09:12 2012] [emerg] (78)Function not implemented: mod_fcgid: Can't create shared memory for size 1192488 bytes security.jail.sysvipc_allowed is enabled (it was needed on 8.x too), nothing else has changed. There are some reports from this, but from earlier versions, and the only confirmed solution was sysvipc_allowed, which works for 8.x, but doesn't with the above version. Any ideas? ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
mpt doesn't propagate read errors and dies on a single sector?
Hi, I have a Sun X4540 with LSI C1068E based SAS controllers (FW version: 1.27.02.00-IT). My problem is if one drive starts to fail with read errors, the machine becomes completely unusable (running stable/9 with ZFS), because -it seems- ZFS can't see that there are read errors on a device, the mpt driver (controller, kernel?) wants to re-issue the operation endlessly. Here is a verbose (dev.mpt.0.debug=7 level) dump: mpt0: Address Reply: SCSI IO Request Reply @ 0xff87ffcfdc00 IOC StatusSuccess IOCLogInfo0x MsgLength 0x09 MsgFlags 0x00 MsgContext0x000200eb Bus: 0 TargetID 3 CDBLength 10 SCSI Status: Check Condition SCSI State: (0x0001)AutoSense_Valid TransferCnt 0x2 SenseCnt 0x0012 ResponseInfo 0x (da3:mpt0:0:3:0): READ(10). CDB: 28 0 3a 38 5d e 0 1 0 0 (da3:mpt0:0:3:0): CAM status: SCSI Status Error (da3:mpt0:0:3:0): SCSI status: Check Condition (da3:mpt0:0:3:0): SCSI sense: MEDIUM ERROR asc:11,0 (Unrecovered read error) (da3:mpt0:0:3:0): Info: 0x3a385d1a (da3:mpt0:0:3:0): Error 5, Unretryable error SCSI IO Request @ 0xff80003046f0 Chain Offset 0x00 MsgFlags 0x00 MsgContext0x000200ea Bus:0 TargetID3 SenseBufferLength 32 LUN: 0x0 Control 0x0200 READ SIMPLEQ DataLength 0x0002 SenseBufAddr0x0c65d5e0 CDB[0:10] 28 00 3a 38 5e 0e 00 01 00 00 SE64 0xff87ffd1c430: Addr=0x00010e858000 FlagsLength=0xd302 64_BIT_ADDRESSING LAST_ELEMENT END_OF_BUFFER END_OF_LIST mpt0: Address Reply: SCSI IO Request Reply @ 0xff87ffcfdd00 IOC StatusSuccess IOCLogInfo0x MsgLength 0x09 MsgFlags 0x00 MsgContext0x000200ea Bus: 0 TargetID 3 CDBLength 10 SCSI Status: Check Condition SCSI State: (0x0001)AutoSense_Valid TransferCnt 0x2 SenseCnt 0x0012 ResponseInfo 0x And I get these check condition SCSI errors endlessly. If ZFS is enabled at boot, the machine can't even start because of this (zpool import never finishes), if I boot without ZFS, and try to import, the zpool command stucks in the vdev_g state: 1163 root 1 200 35440K 5200K vdev_g 6 0:01 0.10% zpool procstat -k 1163 PIDTID COMM TDNAME KSTACK 1163 100116 zpool-mi_switch sleepq_timedwait _sleep biowait vdev_geom_read_guid vdev_geom_open vdev_open vdev_open_children vdev_raidz_open vdev_open vdev_open_children vdev_root_open vdev_open spa_load spa_tryimport zfs_ioc_pool_tryimport zfsdev_ioctl devfs_ioctl_f Could it be that GEOM/ZFS doesn't receive this read error and waits indefinitely for the command to complete? ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Fatal trap 19, Stopped at bge_init_locked+ and bge booting problems
On 02/23/12 21:44, YongHyeon PYUN wrote: I have to ask more information for the controller to Broadcom. Not sure whether I can get some hint at this moment though. :-( Is there anything I can do? I ask this because I have to give back this server very soon. Given that you also have USB related errors, could you completely remove bge(4) in your kernel and see whether it can successfully boot up? I think you can add the following entries to /boot/device.hints without rebuilding kernel. hint.bge.0.disabled="1" hint.bge.1.disabled="1" hint.bge.2.disabled="1" hint.bge.3.disabled="1" This does not help. Removing bge makes it stop here: da0 at ciss0 bus 0 scbus0 target 0 lun 0 da0: Fixed Direct Access SCSI-5 device da0: 135.168MB/s transfers da0: Command Queueing enabled da0: 286070MB (585871964 512 byte sectors: 255H 32S/T 65535C) panic: bootpc_init: no eligible interfaces cpuid = 0 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2a kdb_backtrace() at kdb_backtrace+0x37 panic() at panic+0x187 bootpc_init() at bootpc_init+0x1205 mi_startup() at mi_startup+0x77 btext() at btext+0x2c KDB: enter: panic [ thread pid 0 tid 10 ] Stopped at kdb_enter+0x3b: movq$0,0x976972(%rip) db> Which is completely OK, because there are really no interfaces to boot from. Note that there is no NMI either (maybe because it would happen later in the initialization process). Sadly, I can't boot from disk, but I assume it would work. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Fatal trap 19, Stopped at bge_init_locked+ and bge booting problems
On 02/23/12 05:15, YongHyeon PYUN wrote: bge0: mem 0xf6bf-0xf6bf,0xf6be-0xf6be,0xf6bd-0xf6bd irq 32 at device 0.0 on pci3 bge0: CHIP ID 0x05719001; ASIC REV 0x5719; CHIP REV 0x57190; PCI-E ^^ This controller is new one. Probably BCM5719 A1 but not sure. Yes, it's in a new machine. bge0: Try again This message indicates your controller has ASF/IPMI firmware. Try disabling ASF and see whether it makes any difference. (Change hw.bge.allow_asf tunable to 0). Oh, I always forget that (on the other machines this is set). This is what I get with machdep.panic_on_nmi: 0 machdep.kdb_on_nmi: 0 hw.bge.allow_asf: 0 bge0: mem 0xf6bf-0xf6bf,0xf6be-0xf6be,0xf6bd-0xf6bd irq 32 at device 0.0 on pci3 bge0: CHIP ID 0x05719001; ASIC REV 0x5719; CHIP REV 0x57190; PCI-E bge0: Try again miibus0: on bge0 ukphy0: PHY 1 on miibus0 ukphy0: none, 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, auto, auto-flow bge0: Ethernet address: 3c:4a:92:b2:3c:08 pci0:3:0:1: failed to read VPD data. bge1: mem 0xf6bc-0xf6bc,0xf6bb-0xf6bb,0xf6ba-0xf6ba irq 36 at device 0.1 on pci3 bge1: CHIP ID 0x05719001; ASIC REV 0x5719; CHIP REV 0x57190; PCI-E miibus1: on bge1 brgphy0: PHY 2 on miibus1 brgphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, auto, auto-flow bge1: Ethernet address: 3c:4a:92:b2:3c:09 pci0:3:0:2: failed to read VPD data. bge2: mem 0xf6b9-0xf6b9,0xf6b8-0xf6b8,0xf6b7-0xf6b7 irq 32 at device 0.2 on pci3 bge2: CHIP ID 0x05719001; ASIC REV 0x5719; CHIP REV 0x57190; PCI-E miibus2: on bge2 brgphy1: PHY 3 on miibus2 brgphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, auto, auto-flow bge2: Ethernet address: 3c:4a:92:b2:3c:0a pci0:3:0:3: failed to read VPD data. bge3: mem 0xf6b6-0xf6b6,0xf6b5-0xf6b5,0xf6b4-0xf6b4 irq 36 at device 0.3 on pci3 bge3: CHIP ID 0x05719001; ASIC REV 0x5719; CHIP REV 0x57190; PCI-E miibus3: on bge3 brgphy2: PHY 4 on miibus3 brgphy2: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-master, 1000baseT-FDX, 1000baseT-FDX-master, auto, auto-flow bge3: Ethernet address: 3c:4a:92:b2:3c:0b [...] da0: 286070MB (585871964 512 byte sectors: 255H 32S/T 65535C) NMI ISA 60, EISA ff I/O channel check, likely hardware failure.Sending DHCP Discover packet from interface bge0 (3c:4a:92:b2:3c:08) cd0 at ata3 bus 0 scbus3 target 0 lun 0 cd0: Removable CD-ROM SCSI-0 device cd0: 150.000MB/s transfers (SATA 1.x, UDMA5, ATAPI 12bytes, PIO 8192bytes) cd0: Attempt to query device size failed: NOT READY, Medium not present - tray closed bge0: 11 link states coalesced bge0: link state changed to DOWN ugen0.2: at usbus0 uhub3: on usbus0 bge1: 5 link states coalesced bge1: link state changed to DOWN bge2: link state changed to DOWN bge3: link state changed to DOWN bge0: ugen2.2: at usbus2 uhub4: on usbus2 2 link states coalesced bge0: link state changed to DOWN bge1: 4 link states coalesced bge1: link state changed to DOWN bge0: 4 link states coalesced bge0: link state changed to DOWN Sending DHCP Discover packet from interface bge1 (3c:4a:92:b2:3c:09) uhub3: 6 ports with 6 removable, self powered bge0: usb_alloc_device: set address 2 failed (USB_ERR_TIMEOUT, ignored) 6 link states coalesced bge0: link state changed to DOWN bge1: 2 link states coalesced bge1: link state changed to DOWN Sending DHCP Discover packet from interface bge2 (3c:4a:92:b2:3c:0a) bge0: 2 link states coalesced bge0: link state changed to DOWN bge1: usbd_setup_device_desc: getting device descriptor at addr 2 failed, USB_ERR_TIMEOUT 10 link states coalesced bge1: link state changed to DOWN uhub4: 8 ports with 8 removable, self powered bge0: 4 link states coalesced bge0: link state changed to DOWN bge1: 2 link states coalesced bge1: link state changed to DOWN Sending DHCP Discover packet from interface bge3 (3c:4a:92:b2:3c:0b) bge0: 2 link states coalesced bge0: link state changed to DOWN bge1: 2 link states coalesced bge1: link state changed to DOWN ugen2.3: at usbus2 uhub5: on usbus2 bge0: watchdog timeout -- resetting bge0: link state changed to UP usbd_req_re_enumerate: addr=2, set address failed! (USB_ERR_TIMEOUT, ignored) bge1: 2 link states coalesced bge1: link state changed to DOWN uhub5: 2 ports with 1 removable, self powered bge0: link state changed to DOWN bge1: watchdog timeout -- resetting bge1: usbd_setup_device_desc: getting device descriptor at addr 2 failed, USB_ERR_TIMEOUT 2 link states coalesced bge1: link state changed to DOWN bge0: 2 link states coalesced bge0: link state changed to DOWN ugen1.2: at usbus1 ukbd0: on usbus1 kbd2 at ukbd0 ums0: on usbus1 bge1: 4 link states coalesced bge1: link state changed to DOWN bge0:
Fatal trap 19, Stopped at bge_init_locked+ and bge booting problems
Hi, I get this on a recent stable/9 system with uhci support removed from the kernel config: da0 at ciss0 bus 0 scbus0 target 0 lun 0 da0: Fixed Direct Access SCSI-5 device da0: 135.168MB/s transfers da0: Command Queueing enabled da0: 286070MB (585871964 512 byte sectors: 255H 32S/T 65535C) cd0 at ata3 bus 0 scbus3 target 0 lun 0 cd0: Removable CD-ROM SCSI-0 device cd0: 150.000MB/s transfers (SATA 1.x, UDMA5, ATAPI 12bytes, PIO 8192bytes) cd0: Attempt to query device size failed: NOT READY, Medium not present - tray closed NMI ISA 70, EISA ff I/O channel check, likely hardware failure. Fatal trap 19: non-maskable interrupt trap while in kernel mode cpuid = 0; apic id = 00 instruction pointer = 0x20:0x804543fb stack pointer = 0x28:0x81251e40 frame pointer = 0x28:0x814cf660 code segment= base 0x0, limit 0xf, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags= interrupt enabled, IOPL = 0 current process = 0 (swapper) [ thread pid 0 tid 10 ] Stopped at bge_init_locked+0x233b: movl0x81c(%rsi),%eax db> and this with a plain GENERIC kernel: da0 at ciss0 bus 0 scbus0 target 0 lun 0 da0: Fixed Direct Access SCSI-5 device da0: 135.168MB/s transfers da0: Command Queueing enabled da0: 286070MB (585871964 512 byte sectors: 255H 32S/T 65535C) cd0 at ata3 bus 0 scbus3 target 0 lun 0 cd0: Removable CD-ROM SCSI-0 device cd0: 150.000MB/s transfers (SATA 1.x, UDMA5, ATAPI 12bytes, PIO 8192bytes) cd0: Attempt to query device size failed: NOT READY, Medium not present - tray closed NMI ISA 70, EISA ff I/O channel check, likely hardware failure. Fatal trap 19: non-maskable interrupt trap while in kernel mode cpuid = 0; apic id = 00 instruction pointer = 0x20:0x80711dc5 stack pointer = 0x28:0x81272040 frame pointer = 0x28:0xff907cf44b40 code segment= base 0x0, limit 0xf, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags= interrupt enabled, IOPL = 0 current process = 12 (irq16: uhci0) [ thread pid 12 tid 100098 ] Stopped at uhci_interrupt+0x65:movzwl %ax,%eax db> KDB: stack backtrace: KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2a kdb_backtrace() at kdb_backtrace+0x37 mi_switch() at mi_switch+0x27a turnstile_wait() at turnstile_wait+0x1cb _mtx_lock_sleep() at _mtx_lock_sleep+0xb0 ukbd_poll() at ukbd_poll+0xbe kbdmux_poll() at kbdmux_poll+0x3f sc_cngetc() at sc_cngetc+0xec cncheckc() at cncheckc+0x4a cngetc() at cngetc+0x1c db_readline() at db_readline+0x77 db_read_line() at db_read_line+0x15 db_command_loop() at db_command_loop+0x38 db_trap() at db_trap+0x89 kdb_trap() at kdb_trap+0x101 trap_fatal() at trap_fatal+0x29d trap() at trap+0x10a nmi_calltrap() at nmi_calltrap+0x8 --- trap 0x13, rip = 0x80711dc5, rsp = 0x81272040, rbp = 0xff907cf44b40 --- uhci_interrupt() at uhci_interrupt+0x65 intr_event_execute_handlers() at intr_event_execute_handlers+0x104 ithread_loop() at ithread_loop+0xa4 fork_exit() at fork_exit+0x11f fork_trampoline() at fork_trampoline+0xe --- trap 0, rip = 0, rsp = 0xff907cf44d00, rbp = 0 --- db> After disabling stopping on NMI (kdb_on_nmi), I still can't boot from bge (this is a PXE booted machine), I get this in an infinite loop: bge1: link state changed to DOWN DHCP/BOOTP timeout for server 255.255.255.255 bge1: 3 link states coalesced bge1: link state changed to UP bge0: 2 link states coalesced bge0: link state changed to DOWN bge0: link state changed to UP bge1: link state changed to DOWN bge0: link state changed to DOWN bge0: link state changed to UP bge0: link state changed to DOWN bge1: 2 link states coalesced bge1: link state changed to DOWN bge0: link state changed to UP bge0: link state changed to DOWN bge0: 2 link states coalesced bge0: link state changed to DOWN bge1: 2 link states coalesced bge1: link state changed to DOWN bge0: link state changed to UP bge0: link state changed to DOWN bge0: link state changed to UP bge0: link state changed to DOWN Linux and Windows boot fine on the machine. dmesg up to the point where it crashes: Copyright (c) 1992-2012 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD is a registered trademark of The FreeBSD Foundation. FreeBSD 9.0-STABLE #3: Tue Feb 21 11:57:33 CET 2012 r...@boot.lab:/usr/obj/usr/src/sys/BOOTCLNT amd64 CPU: Intel(R) Xeon(R) CPU E5-2680 0 @ 2.70GHz (2693.57-MHz K8-class CPU) Origin = "GenuineIntel" Id = 0x206d6 Family = 6 Model = 2d Stepping = 6 Features=0xbfebfbff Features2=0x15bee3ff AMD Features=0x2c100800 AMD Features2=0x1 TSC: P-state invariant, performance statistics real memory = 34359738368 (32768 MB) avail memory = 32991268864 (31462 MB) Event timer "LAPIC
Re: Enabling IPSec panics stable/9 (runs OK on stable/8)
On 01/05/12 11:37, VANHULLEBUS Yvan wrote: Strange. may be related to some kind of code optimization As the line juste before is: saidx =&sav->sah->saidx; Could you show the value of&sav->sah->saidx ? And also check if kgdb can print sav->sah->saidx (without the&) ? Oh sorry, the previous console copy was chopped. I've tried sav too: (kgdb) p sav->sah->saidx Variable "sav" is not available. To help you having a quick workaround, do you really need ESP+AH ? Most of the time, people who configure ESP+AH just needs in fact ESP with optional data authentication. I could live without it, you are right. Thanks for the tip. And the crash occurs in some part of the code which deals with encapsulation in encapsulation. This also may explains why I never saw that crash.. Could be. Anyways, it's a permitted option, so it would be good to fix it. I hope I can help you in that somehow. :) Thanks, ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Enabling IPSec panics stable/9 (runs OK on stable/8)
On 01/04/12 17:31, VANHULLEBUS Yvan wrote: On Wed, Jan 04, 2012 at 04:17:41PM +0100, Attila Nagy wrote: [] #7 0x809bf779 in ipsec_process_done (m=0xfe000c7c7a00, isr=0xfe001bf54380) at /data/usr/src/sys/netipsec/ipsec_output.c:170 Here seems to be the problem Can you do the following (in this order) in kgdb: frame 7 p saidx p *saidx (kgdb) frame 7 #7 0x809bf779 in ipsec_process_done (m=0xfe000c7c7a00, isr=0xfe001bf54380) at /data/usr/src/sys/netipsec/ipsec_output.c:170 170 switch (saidx->dst.sa.sa_family) { (kgdb) p saidx No symbol "saidx" in current context. There *is* such a symbol, as confirmed by kgdb's output when you switched to frame 7 ! Could you check that you are running a correct debug kernel ? The kernel config is GENERIC, plus some, so it contains DEBUG=-g makeoptions. This is a limited environment (with a lot of programs missing from the boot image), but I don't think it should affect that. kgdb command line is: ./kgdb kernel.debug vmcore.0 GNU gdb 6.1.1 [FreeBSD] Copyright 2004 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "amd64-marcel-freebsd"... Unread portion of the kernel message buffer: Fatal trap 12: page fault while in kernel mode cpuid = 4; apic id = 04 fault virtual address = 0xa0 fault code = supervisor read data, page not present instruction pointer = 0x20:0x809bf779 stack pointer = 0x28:0xff80002cd350 frame pointer = 0x28:0xff80002cd390 code segment= base 0x0, limit 0xf, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags= interrupt enabled, resume, IOPL = 0 current process = 12 (swi1: netisr 0) trap number = 12 panic: page fault cpuid = 4 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2a kdb_backtrace() at kdb_backtrace+0x37 panic() at panic+0x187 trap_fatal() at trap_fatal+0x290 trap_pfault() at trap_pfault+0x1f9 trap() at trap+0x3df calltrap() at calltrap+0x8 --- trap 0xc, rip = 0x809bf779, rsp = 0xff80002cd350, rbp = 0xff80002cd390 --- ipsec_process_done() at ipsec_process_done+0x119 esp_output_cb() at esp_output_cb+0x1a1 crypto_done() at crypto_done+0x102 swcr_process() at swcr_process+0x1d7 crypto_invoke() at crypto_invoke+0x6b crypto_dispatch() at crypto_dispatch+0xfb esp_output() at esp_output+0x5a2 ipsec4_process_packet() at ipsec4_process_packet+0x1f8 ip_ipsec_output() at ip_ipsec_output+0x16a ip_output() at ip_output+0x526 icmp_reflect() at icmp_reflect+0x339 icmp_input() at icmp_input+0x257 ip_input() at ip_input+0x1de swi_net() at swi_net+0x14d intr_event_execute_handlers() at intr_event_execute_handlers+0x104 ithread_loop() at ithread_loop+0xa4 fork_exit() at fork_exit+0x11f fork_trampoline() at fork_trampoline+0xe --- trap 0, rip = 0, rsp = 0xff80002cdd00, rbp = 0 --- Uptime: 12m32s Dumping 30283 MB (4 chunks) chunk 0: 1MB (139 pages) ... ok chunk 1: 3070MB (785904 pages)panic: bufwrite: buffer is not busy??? cpuid = 4 3054 3038 3022 3006 2990 2974 2958 2942 2926 2910 2894 2878 2862 2846 2830 2814 2798 2782 2766 2750 2734 2718 2702 2686 2670 2654 2638 2622 2606 2590 2574 2558 2542 2526 2510 2494 2478 2462 2446 2430 2414 2398 2382 2366 2350 2334 2318 2302 2286 2270 2254 2238 2206 2190 2174 2158 2142 2126 2110 2094 2078 2062 2046 2030 2014 1998 1982 1966 1950 1934 1918 1902 1886 1870 1854 1838 1822 1806 1790 1774 1758 1742 1726 1710 1694 1678 1662 1646 1630 1614 1598 1582 1566 1550 1534 1518 1502 1486 1470 1454 1438 1422 1406 1390 1374 1358 1342 1326 1310 1294 1278 1262 1246 1230 1214 1198 1182 1166 1150 1134 1118 1102 1086 1070 1054 1038 1022 1006 990 974 958 942 926 910 894 878 862 846 830 814 798 782 766 750 734 718 702 686 670 654 638 622 606 590 574 558 542 526 510 494 478 462 446 430 414 398 382 366 350 334 318 302 286 270 254 238 222 206 190 174 158 142 126 110 94 78 62 46 30 14 ... ok chunk 2: 1MB (256 pages) ... ok chunk 3: 27212MB (6966272 pages) 27197 27181 27165 27149 27133 27117 27101 27085 27069 27053 27037 27021 27005 26989 26973 26957 26941 26925 26909 26893 26877 26861 26845 26829 26813 26797 26781 26765 26749 26733 26717 26701 26685 26669 26653 26637 26621 26605 26589 26573 26557 26541 26525 26509 26493 26477 26461 26445 26429 26413 26397 26381 26365 26349 26333 26317 26301 26285 26269 26253 26237 26221 26205 26189 26173 26157 26141 26125 26109 26093 26077 26061 26045 26029 26013 25997 25981 25965 25949 25933 25917 25901 25885 25869 25853 25837 25821 25805 25789 25773
Re: Enabling IPSec panics stable/9 (runs OK on stable/8)
Hi, On 01/04/12 15:51, VANHULLEBUS Yvan wrote: I've just upgraded a 8-STABLE box to 9-STABLE (well, just few commits before it has been tagged as STABLE), which runs from NFS (pxebooted). It has some IPSec config in ipsec.conf, like this for several boxes: add 172.28.16.4 172.16.248.2 ah 15704 -A hmac-md5 "asdfgh"; add 172.16.248.2 172.28.16.4 ah 24504 -A hmac-md5 "asdfgh"; add 172.28.16.4 172.16.248.2 esp 15705 -E blowfish-cbc "hgfdsa"; add 172.16.248.2 172.28.16.4 esp 24505 -E blowfish-cbc "hgfdsa"; spdadd 172.28.16.4 172.16.248.2 any -P out ipsec esp/transport/172.28.16.4-172.16.248.2/default ah/transport/172.28.16.4-172.16.248.2/default; There is probably nothing related to the crash, but do you really use static IPsec without IKE keying Yes. :) It runs on an intranet, but there's a need to encrypt traffic. [] kgdb says: (kgdb) bt #0 doadump (textdump=1) at /data/usr/src/sys/kern/kern_shutdown.c:260 #1 0x80845705 in kern_reboot (howto=260) at /data/usr/src/sys/kern/kern_shutdown.c:442 #2 0x80845bb1 in panic (fmt=Variable "fmt" is not available. ) at /data/usr/src/sys/kern/kern_shutdown.c:607 #3 0x80b167a0 in trap_fatal (frame=0xc, eva=Variable "eva" is not available. ) at /data/usr/src/sys/amd64/amd64/trap.c:819 #4 0x80b16ae9 in trap_pfault (frame=0xff80002cd2a0, usermode=0) at /data/usr/src/sys/amd64/amd64/trap.c:735 #5 0x80b16faf in trap (frame=0xff80002cd2a0) at /data/usr/src/sys/amd64/amd64/trap.c:474 #6 0x80b012ef in calltrap () at /data/usr/src/sys/amd64/amd64/exception.S:228 #7 0x809bf779 in ipsec_process_done (m=0xfe000c7c7a00, isr=0xfe001bf54380) at /data/usr/src/sys/netipsec/ipsec_output.c:170 Here seems to be the problem Can you do the following (in this order) in kgdb: frame 7 p saidx p *saidx (kgdb) frame 7 #7 0x809bf779 in ipsec_process_done (m=0xfe000c7c7a00, isr=0xfe001bf54380) at /data/usr/src/sys/netipsec/ipsec_output.c:170 170 switch (saidx->dst.sa.sa_family) { (kgdb) p saidx No symbol "saidx" in current context. The latest will probably generate an error, as (if you have the exact same ipsec_output.c as I have from HEAD) saidx will probably have an invalid adress. I have the same as in HEAD. [...] 8-STABLE runs fine with the same config. Strange I'll review changes in IPsec stack which have been done in STABLE/9 and not backported to STABLE/8. Oh, sorry, not quite an up-to-date 8-STABLE, it's from Sat May 21 22:05:26 CEST 2011 (csup'd some hours earlier). Should I check with a more recent version? Does that help? Thanks for helping. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Enabling IPSec panics stable/9 (runs OK on stable/8)
Hi, I've just upgraded a 8-STABLE box to 9-STABLE (well, just few commits before it has been tagged as STABLE), which runs from NFS (pxebooted). It has some IPSec config in ipsec.conf, like this for several boxes: add 172.28.16.4 172.16.248.2 ah 15704 -A hmac-md5 "asdfgh"; add 172.16.248.2 172.28.16.4 ah 24504 -A hmac-md5 "asdfgh"; add 172.28.16.4 172.16.248.2 esp 15705 -E blowfish-cbc "hgfdsa"; add 172.16.248.2 172.28.16.4 esp 24505 -E blowfish-cbc "hgfdsa"; spdadd 172.28.16.4 172.16.248.2 any -P out ipsec esp/transport/172.28.16.4-172.16.248.2/default ah/transport/172.28.16.4-172.16.248.2/default; Running /etc/rc.d/ipsec start instantly panics it with: Fatal trap 12: page fault while in kernel mode cpuid = 1; apic id = 01 fault virtual address = 0xa0 fault code = supervisor read data, page not present instruction pointer = 0x20:0x809bf779 stack pointer = 0x28:0xff80002cd350 frame pointer = 0x28:0xff80002cd390 code segment= base 0x0, limit 0xf, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags= interrupt enabled, resume, IOPL = 0 current process = 12 (swi1: netisr 0) trap number = 12 panic: page fault cpuid = 1 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2a kdb_backtrace() at kdb_backtrace+0x37 panic() at panic+0x187 trap_fatal() at trap_fatal+0x290 trap_pfault() at trap_pfault+0x1f9 trap() at trap+0x3df calltrap() at calltrap+0x8 --- trap 0xc, rip = 0x809bf779, rsp = 0xff80002cd350, rbp = 0xff80002cd390 --- ipsec_process_done() at ipsec_process_done+0x119 esp_output_cb() at esp_output_cb+0x1a1 crypto_done() at crypto_done+0x102 swcr_process() at swcr_process+0x1d7 crypto_invoke() at crypto_invoke+0x6b crypto_dispatch() at crypto_dispatch+0xfb esp_output() at esp_output+0x5a2 ipsec4_process_packet() at ipsec4_process_packet+0x1f8 ip_ipsec_output() at ip_ipsec_output+0x16a ip_output() at ip_output+0x526 icmp_reflect() at icmp_reflect+0x339 icmp_input() at icmp_input+0x257 ip_input() at ip_input+0x1de swi_net() at swi_net+0x14d intr_event_execute_handlers() at intr_event_execute_handlers+0x104 ithread_loop() at ithread_loop+0xa4 fork_exit() at fork_exit+0x11f fork_trampoline() at fork_trampoline+0xe --- trap 0, rip = 0, rsp = 0xff80002cdd00, rbp = 0 --- Uptime: 3m31s The machine in question is a VMWare virtual environment (if it counts), the kernel config is GENERIC plus this: options BOOTP options BOOTP_NFSV3 options BOOTP_NFSROOT options NFSCLIENT device carp device crypto device cryptodev options IPSEC options MAC options ROUTETABLES=1 options KDB options DDB options KDB_UNATTENDED kgdb says: (kgdb) bt #0 doadump (textdump=1) at /data/usr/src/sys/kern/kern_shutdown.c:260 #1 0x80845705 in kern_reboot (howto=260) at /data/usr/src/sys/kern/kern_shutdown.c:442 #2 0x80845bb1 in panic (fmt=Variable "fmt" is not available. ) at /data/usr/src/sys/kern/kern_shutdown.c:607 #3 0x80b167a0 in trap_fatal (frame=0xc, eva=Variable "eva" is not available. ) at /data/usr/src/sys/amd64/amd64/trap.c:819 #4 0x80b16ae9 in trap_pfault (frame=0xff80002cd2a0, usermode=0) at /data/usr/src/sys/amd64/amd64/trap.c:735 #5 0x80b16faf in trap (frame=0xff80002cd2a0) at /data/usr/src/sys/amd64/amd64/trap.c:474 #6 0x80b012ef in calltrap () at /data/usr/src/sys/amd64/amd64/exception.S:228 #7 0x809bf779 in ipsec_process_done (m=0xfe000c7c7a00, isr=0xfe001bf54380) at /data/usr/src/sys/netipsec/ipsec_output.c:170 #8 0x809ce931 in esp_output_cb (crp=0xfe011103c058) at /data/usr/src/sys/netipsec/xform_esp.c:1007 #9 0x809f4e12 in crypto_done (crp=0xfe011103c058) at /data/usr/src/sys/opencrypto/crypto.c:1156 #10 0x809f89c7 in swcr_process (dev=Variable "dev" is not available. ) at /data/usr/src/sys/opencrypto/cryptosoft.c:1054 #11 0x809f5c9b in crypto_invoke (cap=0xfe000c12f700, crp=0xfe011103c058, hint=0) at cryptodev_if.h:53 #12 0x809f6acb in crypto_dispatch (crp=0xfe011103c058) at /data/usr/src/sys/opencrypto/crypto.c:806 #13 0x809cef82 in esp_output (m=0xfe000c7c7a00, isr=0xfe001bf54380, mp=Variable "mp" is not available. ) at /data/usr/src/sys/netipsec/xform_esp.c:907 #14 0x809bfa98 in ipsec4_process_packet (m=0xfe000c7c7a00, isr=0xfe001bf54380, flags=Variable "flags" is not available. ) at /data/usr/src/sys/netipsec/ipsec_output.c:580 #15 0x8096f5da in ip_ips
Re: tmpfs is zero bytes (no free space), maybe a zfs bug?
On 02/10/2011 05:56 PM, Bruce Cran wrote: On Wed, 19 Jan 2011 11:09:31 +0100 Attila Nagy wrote: On 01/19/11 09:46, Jeremy Chadwick wrote: On Wed, Jan 19, 2011 at 09:37:35AM +0100, Attila Nagy wrote: I first noticed this problem on machines with more memory (32GB eg.), but now it happens on 4G machines too: tmpfs 0B 0B 0B 100%/tmp FreeBSD builder 8.2-PRERELEASE FreeBSD 8.2-PRERELEASE #0: Sat Jan 8 22:11:54 CET 2011 Maybe it's related, that I use zfs on these machines... Sometimes it grows and shrinks, but generally there is no space even for a small file, or a socket to create. http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060867.html Oh crap. :( I hope somebody can find the time to look into this, it's pretty annoying... It's also listed as a bug on OpenSolaris: http://bugs.opensolaris.org/bugdatabase/view_bug.do;?bug_id=6804661 ZFS is a great innovation, which forces sysadmins to learn kernel and VM internals. :-O ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: tmpfs is zero bytes (no free space), maybe a zfs bug?
On 01/30/11 12:09, Kostik Belousov wrote: On Wed, Jan 19, 2011 at 05:27:38PM +0100, Ivan Voras wrote: On 19 January 2011 16:02, Kostik Belousov wrote: http://people.freebsd.org/~ivoras/diffs/tmpfs.h.patch I don't think this is a complete solution but it's a start. If you can, try it and see if it helps. This is not a start, and actually a step in the wrong direction. Tmpfs is wrong now, but the patch would make the wrongness even bigger. Issue is that the current tmpfs calculation should not depend on the length of the inactive queue or the amount of free pages. This data only measures the pressure on the pagedaemon, and has absolutely no relation to the amount of data that can be put into anonymous objects before the system comes out of swap. vm_lowmem handler is invoked in two situations: - when KVA cannot satisfy the request for the space allocation; - when pagedaemon have to start the scan. None of the situations has any direct correlation with the fact that tmpfs needs to check, that is "Is there enough swap to keep all my future anonymous memory requests ?". Might be, swap reservation numbers can be useful to the tmpfs reporting. Also might be, tmpfs should reserve the swap explicitely on start, instead of making attempts to guess how much can be allocated at random moment. Thank you for your explanation! I'm still not very familiar with VM and VFS. Could you also read my report at http://www.mail-archive.com/freebsd-current@freebsd.org/msg126491.html ? I'm curious about the fact that there is lots of 'free' memory here in the same situation. This is another ugliness in the dynamic calculation. Your wired is around 15GB, that is always greater then available swap + free + inactive. As result, tmpfs_mem_info() always returns 0. In this situation TMPFS_PAGES_MAX() seems to return negative value, and then TMPFS_PAGES_AVAIL() clamps at 0. Well, if nobody can take care of this now, could you please state this in the BUGS section of the tmpfs man page? Thanks, ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: tmpfs is zero bytes (no free space), maybe a zfs bug?
On 01/19/11 09:46, Jeremy Chadwick wrote: On Wed, Jan 19, 2011 at 09:37:35AM +0100, Attila Nagy wrote: I first noticed this problem on machines with more memory (32GB eg.), but now it happens on 4G machines too: tmpfs 0B 0B 0B 100%/tmp FreeBSD builder 8.2-PRERELEASE FreeBSD 8.2-PRERELEASE #0: Sat Jan 8 22:11:54 CET 2011 Maybe it's related, that I use zfs on these machines... Sometimes it grows and shrinks, but generally there is no space even for a small file, or a socket to create. http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/060867.html Oh crap. :( I hope somebody can find the time to look into this, it's pretty annoying... ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
tmpfs is zero bytes (no free space), maybe a zfs bug?
Hi, I first noticed this problem on machines with more memory (32GB eg.), but now it happens on 4G machines too: tmpfs 0B 0B 0B 100% /tmp FreeBSD builder 8.2-PRERELEASE FreeBSD 8.2-PRERELEASE #0: Sat Jan 8 22:11:54 CET 2011 Maybe it's related, that I use zfs on these machines... Sometimes it grows and shrinks, but generally there is no space even for a small file, or a socket to create. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 12/16/2010 01:44 PM, Martin Matuska wrote: Hi everyone, following the announcement of Pawel Jakub Dawidek (p...@freebsd.org) I am providing a ZFSv28 testing patch for 8-STABLE. Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz Link to mfsBSD ISO files for testing (i386 and amd64): http://mfsbsd.vx.sk/iso/zfs-v28/8.2-beta-zfsv28-amd64.iso http://mfsbsd.vx.sk/iso/zfs-v28/8.2-beta-zfsv28-i386.iso The root password for the ISO files: "mfsroot" The ISO files work on real systems and in virtualbox. They conatin a full install of FreeBSD 8.2-PRERELEASE with ZFS v28, simply use the provided "zfsinstall" script. The patch is against FreeBSD 8-STABLE as of 2010-12-15. When applying the patch be sure to use correct options for patch(1) and make sure the file sys/cddl/compat/opensolaris/sys/sysmacros.h gets deleted: # cd /usr/src # fetch http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz # xz -d stable-8-zfsv28-20101215.patch.xz # patch -E -p0< stable-8-zfsv28-20101215.patch # rm sys/cddl/compat/opensolaris/sys/sysmacros.h I've just got a panic: http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/IMAGE_006.jpg The panic line for google: panic: solaris assert: task->ost_magic == TASKQ_MAGIC, file: /usr/src/sys/modules/zfs/../../cddl/compat/opensolaris/kern/opensolaris_taskq.c, line: 150 I hope this is enough for debugging, if it's not yet otherwise known. If not, I will try to catch it againt and make a dump. Thanks, ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/10/2011 09:57 AM, Pawel Jakub Dawidek wrote: On Sun, Jan 09, 2011 at 12:52:56PM +0100, Attila Nagy wrote: [...] I've finally found the time to read the v28 patch and figured out the problem: vfs.zfs.l2arc_noprefetch was changed to 1, so it doesn't use the prefetched data on the L2ARC devices. This is a major hit in my case. Enabling this again restored the previous hit rates and lowered the load on the hard disks significantly. Well, not storing prefetched data on L2ARC vdevs is the default is Solaris. For some reason it was changed by kmacy@ in r205231. Not sure why and we can't ask him now, I'm afraid. I just sent an e-mail to What happened to him? Brendan Gregg from Oracle who originally implemented L2ARC in ZFS why this is turned off by default. Once I get answer we can think about turning it on again. I think it makes some sense as a stupid form of preferring random IO in the L2ARC instead of sequential. But if I rely on auto tuning and let prefetch enabled, even a busy mailserver will prefetch a lot of blocks and I think that's a fine example of random IO (also, it makes the system unusable, but that's another story). Having this choice is good, and in this case enabling this makes sense for me. I don't know any reasons about why you wouldn't use all of your L2ARC space (apart from sparing the quickly wearing out flash space and move disk heads instead), but I'm sure Brendan made this choice with a good reason. If you get an answer, please tell us. :) Thanks, ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/10/2011 10:02 AM, Pawel Jakub Dawidek wrote: On Sun, Jan 09, 2011 at 12:49:27PM +0100, Attila Nagy wrote: No, it's not related. One of the disks in the RAIDZ2 pool went bad: (da4:arcmsr0:0:4:0): READ(6). CDB: 8 0 2 10 10 0 (da4:arcmsr0:0:4:0): CAM status: SCSI Status Error (da4:arcmsr0:0:4:0): SCSI status: Check Condition (da4:arcmsr0:0:4:0): SCSI sense: MEDIUM ERROR asc:11,0 (Unrecovered read error) and it seems it froze the whole zpool. Removing the disk by hand solved the problem. I've seen this previously on other machines with ciss. I wonder why ZFS didn't throw it out of the pool. Such hangs happen when I/O never returns. ZFS doesn't timeout I/O requests on its own, this is driver's responsibility. It is still strange that the driver didn't pass I/O error up to ZFS or it might as well be ZFS bug, but I don't think so. Indeed, it may to be a controller/driver bug. The newly released (last december) firmware says something about a similar problem. I've upgraded, we'll see whether it will help next time a drive goes awry. I've only seen these errors in dmesg, not in zpool status, there everything was clear (all zeroes). BTW, I've swapped those bad drives (da4, which reported the above errors, and da16, which didn't reported anything to the OS, it was just plain bad according to the controller firmware -and after its deletion, I could offline da4, so it seems it's the real cause, see my previous e-mail), and zpool replaced first da4, but after some seconds of thinking all IO on all disks deceased. After waiting some minutes, it was still the same, so I've rebooted. Then I noticed that a scrub is going on, so I stopped it. Then the zpool replace da4 went fine, it started to resilver the disk. But another zpool replace (for da16) causes the same error: some seconds of IO, then nothing and it stuck in that. Has anybody tried replacing two drives simultaneously with the zfs v28 patch? (this is a stripe of two raidz2s and da4 and da16 are in different raidz2) ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/09/2011 01:18 PM, Jeremy Chadwick wrote: On Sun, Jan 09, 2011 at 12:49:27PM +0100, Attila Nagy wrote: On 01/09/2011 10:00 AM, Attila Nagy wrote: On 12/16/2010 01:44 PM, Martin Matuska wrote: Hi everyone, following the announcement of Pawel Jakub Dawidek (p...@freebsd.org) I am providing a ZFSv28 testing patch for 8-STABLE. Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've got an IO hang with dedup enabled (not sure it's related, I've started to rewrite all data on pool, which makes a heavy load): The processes are in various states: 65747 1001 1 54 10 28620K 24360K tx->tx 0 6:58 0.00% cvsup 80383 1001 1 54 10 40616K 30196K select 1 5:38 0.00% rsync 1501 www 1 440 7304K 2504K zio->i 0 2:09 0.00% nginx 1479 www 1 440 7304K 2416K zio->i 1 2:03 0.00% nginx 1477 www 1 440 7304K 2664K zio->i 0 2:02 0.00% nginx 1487 www 1 440 7304K 2376K zio->i 0 1:40 0.00% nginx 1490 www 1 440 7304K 1852K zfs 0 1:30 0.00% nginx 1486 www 1 440 7304K 2400K zfsvfs 1 1:05 0.00% nginx And everything which wants to touch the pool is/becomes dead. Procstat says about one process: # procstat -k 1497 PIDTID COMM TDNAME KSTACK 1497 100257 nginx-mi_switch sleepq_wait __lockmgr_args vop_stdlock VOP_LOCK1_APV null_lock VOP_LOCK1_APV _vn_lock nullfs_root lookup namei vn_open_cred kern_openat syscallenter syscall Xfast_syscall No, it's not related. One of the disks in the RAIDZ2 pool went bad: (da4:arcmsr0:0:4:0): READ(6). CDB: 8 0 2 10 10 0 (da4:arcmsr0:0:4:0): CAM status: SCSI Status Error (da4:arcmsr0:0:4:0): SCSI status: Check Condition (da4:arcmsr0:0:4:0): SCSI sense: MEDIUM ERROR asc:11,0 (Unrecovered read error) and it seems it froze the whole zpool. Removing the disk by hand solved the problem. I've seen this previously on other machines with ciss. I wonder why ZFS didn't throw it out of the pool. Hold on a minute. An unrecoverable read error does not necessarily mean the drive is bad, it could mean that the individual LBA that was attempted to be read resulted in ASC 0x11 (MEDIUM ERROR) (e.g. a bad block was encountered). I would check SMART stats on the disk (since these are probably SATA given use of arcmsr(4)) and provide those. *That* will tell you if the disk is bad. I'll help you decode the attributes values if you provide them. You are right, and I gave incorrect information. There are a lot more errors for that disk in the logs, and the zpool was frozen. I tried to offline the given disk. That helped in the ciss case, where the symptom is the same, or something similar, like there is no IO for ages, then something small and nothing for long seconds/minutes, and there are no errors logged. zpool status reported no errors, and the dmesg was clear too. There I could find the bad disk by watching gstat output and there I saw when the very small amount of IO was done, there was one disk with response times well above a second, while the others responded quickly. There the zpool offline helped. Here not, the command just got hang, like everything else. So what I did then: got into the areca-cli and searched for errors. One disk was set to failed and it seemed to be the cause. I've removed it (and did a camcontrol rescan, but I'm not sure it was necessary or not), and suddenly the zpool offline finished and everything went back to normal. But there are two controllers in the system and now I see that the above disk is on ctrl 1, while the one I have removed is on ctrl 2. I was misleaded by their same position. So now I have an offlined disk (which produces read errors, but I couldn't see them in the zpool output) and another, which is shown as failed in the RAID controller and got removed by hand (and solved the situation): NAME STATE READ WRITE CKSUM data DEGRADED 0 0 0 raidz2-0 DEGRADED 0 0 0 label/disk20-01 ONLINE 0 0 0 label/disk20-02 ONLINE 0 0 0 label/disk20-03 ONLINE 0 0 0 label/disk20-04 ONLINE 0 0 0 label/disk20-05 OFFLINE 0 0 0 label/disk20-06 ONLINE 0 0 0 label/disk20-07 ONLINE 0 0 0 label/disk20-08 ONLINE 0 0 0 label/disk20-09 ONLINE 0 0 0 label/disk20-10 ONLINE 0 0 0 label/disk20-11 ONLINE 0 0 0 label/disk20-12 ONLINE 0 0 0 raidz2-1 DEGRADED 0 0 0 label/disk21-01 ONLINE 0 0 0 label/disk21-02 ONLINE 0 0 0 label/disk21-03 ONLINE
Re: New ZFSv28 patchset for 8-STABLE
On 01/01/2011 08:09 PM, Artem Belevich wrote: On Sat, Jan 1, 2011 at 10:18 AM, Attila Nagy wrote: What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) ... Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. The fact that L2 ARC is full does not mean that it contains the right data. Initial L2ARC warm up happens at a much higher rate than the rate L2ARC is updated after it's been filled initially. Even accelerated warm-up took almost a day in your case. In order for L2ARC to warm up properly you may have to wait quite a bit longer. My guess is that it should slowly improve over the next few days as data goes through L2ARC and those bits that are hit more often take residence there. The larger your data set, the longer it will take for L2ARC to catch the right data. Do you have similar graphs from pre-patch system just after reboot? I suspect that it may show similarly abysmal L2ARC hit rates initially, too. I've finally found the time to read the v28 patch and figured out the problem: vfs.zfs.l2arc_noprefetch was changed to 1, so it doesn't use the prefetched data on the L2ARC devices. This is a major hit in my case. Enabling this again restored the previous hit rates and lowered the load on the hard disks significantly. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/09/2011 10:00 AM, Attila Nagy wrote: On 12/16/2010 01:44 PM, Martin Matuska wrote: Hi everyone, following the announcement of Pawel Jakub Dawidek (p...@freebsd.org) I am providing a ZFSv28 testing patch for 8-STABLE. Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've got an IO hang with dedup enabled (not sure it's related, I've started to rewrite all data on pool, which makes a heavy load): The processes are in various states: 65747 1001 1 54 10 28620K 24360K tx->tx 0 6:58 0.00% cvsup 80383 1001 1 54 10 40616K 30196K select 1 5:38 0.00% rsync 1501 www 1 440 7304K 2504K zio->i 0 2:09 0.00% nginx 1479 www 1 440 7304K 2416K zio->i 1 2:03 0.00% nginx 1477 www 1 440 7304K 2664K zio->i 0 2:02 0.00% nginx 1487 www 1 440 7304K 2376K zio->i 0 1:40 0.00% nginx 1490 www 1 440 7304K 1852K zfs 0 1:30 0.00% nginx 1486 www 1 440 7304K 2400K zfsvfs 1 1:05 0.00% nginx And everything which wants to touch the pool is/becomes dead. Procstat says about one process: # procstat -k 1497 PIDTID COMM TDNAME KSTACK 1497 100257 nginx-mi_switch sleepq_wait __lockmgr_args vop_stdlock VOP_LOCK1_APV null_lock VOP_LOCK1_APV _vn_lock nullfs_root lookup namei vn_open_cred kern_openat syscallenter syscall Xfast_syscall No, it's not related. One of the disks in the RAIDZ2 pool went bad: (da4:arcmsr0:0:4:0): READ(6). CDB: 8 0 2 10 10 0 (da4:arcmsr0:0:4:0): CAM status: SCSI Status Error (da4:arcmsr0:0:4:0): SCSI status: Check Condition (da4:arcmsr0:0:4:0): SCSI sense: MEDIUM ERROR asc:11,0 (Unrecovered read error) and it seems it froze the whole zpool. Removing the disk by hand solved the problem. I've seen this previously on other machines with ciss. I wonder why ZFS didn't throw it out of the pool. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 12/16/2010 01:44 PM, Martin Matuska wrote: Hi everyone, following the announcement of Pawel Jakub Dawidek (p...@freebsd.org) I am providing a ZFSv28 testing patch for 8-STABLE. Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've got an IO hang with dedup enabled (not sure it's related, I've started to rewrite all data on pool, which makes a heavy load): The processes are in various states: 65747 1001 1 54 10 28620K 24360K tx->tx 0 6:58 0.00% cvsup 80383 1001 1 54 10 40616K 30196K select 1 5:38 0.00% rsync 1501 www 1 440 7304K 2504K zio->i 0 2:09 0.00% nginx 1479 www 1 440 7304K 2416K zio->i 1 2:03 0.00% nginx 1477 www 1 440 7304K 2664K zio->i 0 2:02 0.00% nginx 1487 www 1 440 7304K 2376K zio->i 0 1:40 0.00% nginx 1490 www 1 440 7304K 1852K zfs 0 1:30 0.00% nginx 1486 www 1 440 7304K 2400K zfsvfs 1 1:05 0.00% nginx And everything which wants to touch the pool is/becomes dead. Procstat says about one process: # procstat -k 1497 PIDTID COMM TDNAME KSTACK 1497 100257 nginx-mi_switch sleepq_wait __lockmgr_args vop_stdlock VOP_LOCK1_APV null_lock VOP_LOCK1_APV _vn_lock nullfs_root lookup namei vn_open_cred kern_openat syscallenter syscall Xfast_syscall ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/03/2011 10:35 PM, Bob Friesenhahn wrote: After four days, the L2 hit rate is still hovering around 10-20 percents (was between 60-90), so I think it's clearly a regression in the ZFSv28 patch... And the massive growth in CPU usage can also very nicely be seen... I've updated the graphs at (switch time can be checked on the zfs-mem graph): http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/ There is a new phenomenom: the large IOPS peaks. I use this munin script on a lot of machines and never seen anything like this... I'm not sure whether it's related or not. It is not so clear that there is a problem. I am not sure what you are using this server for but it is wise The IO pattern has changed radically, so for me it's a problem. to consider that this is the funny time when a new year starts, SPAM delivery goes through the roof, and employees and customers behave differently. You chose the worst time of the year to implement the change and observe behavior. It's a free software mirror, ftp.fsn.hu, and I'm sure that it's (the very low hit rate and the increased CPU usage) not related to the time when I made the switch. CPU use is indeed increased somewhat. A lower loading of the l2arc is not necessarily a problem. The l2arc is usually bandwidth limited compared with main store so if bulk data can not be cached in RAM, then it is best left in main store. A smarter l2arc algorithm could put only the data producing the expensive IOPS (the ones requiring a seek) in the l2arc, lessening the amount of data cached on the device. That would make sense, if I wouldn't have 100-120 IOPS (for 7k2 RPM disks, it's about their max, gstat tells me the same) on the disks, and as low as 10 percents of L2 hit rate. What's smarter? Having 60-90% hit rate from the SSDs and moving the slow disk heads less, or having 10-20 percent of hit rate and kill the disks with random IO? If you are right, ZFS tries to be too smart and falls on its face with this kind of workload. BTW, I've checked the v15-v28 patch for arc.c, and I can't see any L2ARC related change there. I'm not sure whether the hypothetical logic would be there, or a different file, I haven't read it end to end. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/01/2011 08:09 PM, Artem Belevich wrote: On Sat, Jan 1, 2011 at 10:18 AM, Attila Nagy wrote: What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) ... Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. The fact that L2 ARC is full does not mean that it contains the right data. Initial L2ARC warm up happens at a much higher rate than the rate L2ARC is updated after it's been filled initially. Even accelerated warm-up took almost a day in your case. In order for L2ARC to warm up properly you may have to wait quite a bit longer. My guess is that it should slowly improve over the next few days as data goes through L2ARC and those bits that are hit more often take residence there. The larger your data set, the longer it will take for L2ARC to catch the right data. Do you have similar graphs from pre-patch system just after reboot? I suspect that it may show similarly abysmal L2ARC hit rates initially, too. After four days, the L2 hit rate is still hovering around 10-20 percents (was between 60-90), so I think it's clearly a regression in the ZFSv28 patch... And the massive growth in CPU usage can also very nicely be seen... I've updated the graphs at (switch time can be checked on the zfs-mem graph): http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/ There is a new phenomenom: the large IOPS peaks. I use this munin script on a lot of machines and never seen anything like this... I'm not sure whether it's related or not. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/02/2011 05:06 AM, J. Hellenthal wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 01/01/2011 13:18, Attila Nagy wrote: On 12/16/2010 01:44 PM, Martin Matuska wrote: Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've used this: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101223-nopython.patch.xz on a server with amd64, 8 G RAM, acting as a file server on ftp/http/rsync, the content being read only mounted with nullfs in jails, and the daemons use sendfile (ftp and http). The effects can be seen here: http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/ the exact moment of the switch can be seen on zfs_mem-week.png, where the L2 ARC has been discarded. What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) Maybe I could accept the higher system load as normal, because there were a lot of things changed between v15 and v28 (but I was hoping if I use the same feature set, it will require less CPU), but dropping the L2ARC hit rate so radically seems to be a major issue somewhere. As you can see from the memory stats, I have enough kernel memory to hold the L2 headers, so the L2 devices got filled up to their maximum capacity. Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. Running arc_summary.pl[1] -p4 should print a summary about your l2arc and you should also notice in that section that there is a high number of "SPA Mismatch" mine usually grew to around 172k before I would notice a crash and I could reliably trigger this while in scrub. What ever is causing this needs desperate attention! I emailed mm@ privately off-list when I noticed this going on but have not received any feedback as of yet. It's at zero currently (2 days of uptime): kstat.zfs.misc.arcstats.l2_write_spa_mismatch: 0 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 01/01/2011 08:09 PM, Artem Belevich wrote: On Sat, Jan 1, 2011 at 10:18 AM, Attila Nagy wrote: What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) ... Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. The fact that L2 ARC is full does not mean that it contains the right data. Initial L2ARC warm up happens at a much higher rate than the rate L2ARC is updated after it's been filled initially. Even accelerated warm-up took almost a day in your case. In order for L2ARC to warm up properly you may have to wait quite a bit longer. My guess is that it should slowly improve over the next few days as data goes through L2ARC and those bits that are hit more often take residence there. The larger your data set, the longer it will take for L2ARC to catch the right data. Do you have similar graphs from pre-patch system just after reboot? I suspect that it may show similarly abysmal L2ARC hit rates initially, too. Sadly no, but I remember that I've seen increasing hit rates as the cache grew, that's what I wrote the email after one and a half days. Currently it's at the same level, when it was right after the reboot... We'll see after few days. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: New ZFSv28 patchset for 8-STABLE
On 12/16/2010 01:44 PM, Martin Matuska wrote: Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've used this: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101223-nopython.patch.xz on a server with amd64, 8 G RAM, acting as a file server on ftp/http/rsync, the content being read only mounted with nullfs in jails, and the daemons use sendfile (ftp and http). The effects can be seen here: http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/ the exact moment of the switch can be seen on zfs_mem-week.png, where the L2 ARC has been discarded. What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) Maybe I could accept the higher system load as normal, because there were a lot of things changed between v15 and v28 (but I was hoping if I use the same feature set, it will require less CPU), but dropping the L2ARC hit rate so radically seems to be a major issue somewhere. As you can see from the memory stats, I have enough kernel memory to hold the L2 headers, so the L2 devices got filled up to their maximum capacity. Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. Thanks, ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Pyun YongHyeon wrote: > On Tue, Mar 30, 2010 at 05:57:45PM +0200, Attila Nagy wrote: > >> Jonathan Feally wrote: >> >>> Attila Nagy wrote: >>> >>>>>>>> Bingo, this solved the problem. The current uptime nears four days. >>>>>>>> Previously I couldn't go further than a day. >>>>>>>> >>>>>>>> The machine gets very light TCP load (and other machines which >>>>>>>> get work >>>>>>>> well), so I guess it's UDP RX or TX checksum related >>>>>>>> >>>>>>>> >>> I also have had my network go dead on a recent 8.0-STABLE on bge >>> system. Console is alive, but network just stops. I am running it as a >>> router with untagged on bge0 and nat of traffic on vlan201 tagged on >>> top of bge1. I haven't had it lock up in 3 days, but I will try the >>> -txcsum and -rxcsum on both interfaces to see if the problem still >>> persists or not. I do have a lot of tcp traffic, but there is also >>> unsolicited udp flying in as well. >>> >> Well, it's a short time to judge from, but with rx,txcsum disabled, the >> machine froze nearly instantly (less than one hour of uptime), while >> with tso disabled, it still works. >> So for now I think tso causes the problems. >> BTW, now that we are talking about that, I remember that I've disabled >> it on a lot of machines previously, because I've had strange issues. >> > > Would you show me the dmesg output(only bce(4) part)? > Sure: bce0: mem 0xfa00-0xfbff irq 16 at device 0.0 on pci7 miibus0: on bce0 brgphy0: PHY 2 on miibus0 brgphy0: 1000baseSX-FDX, 2500baseSX-FDX, auto bce0: Ethernet address: 00:1b:78:75:f0:34 bce0: [ITHREAD] bce0: ASIC (0x57081021); Rev (B2); Bus (PCI-X, 64-bit, 133MHz); B/C (4.4.1); Flags (MSI|2.5G) bce1: mem 0xf600-0xf7ff irq 16 at device 0.0 on pci3 miibus1: on bce1 brgphy1: PHY 2 on miibus1 brgphy1: 1000baseSX-FDX, 2500baseSX-FDX, auto bce1: Ethernet address: 00:1b:78:75:f0:38 bce1: [ITHREAD] bce1: ASIC (0x57081021); Rev (B2); Bus (PCI-X, 64-bit, 133MHz); B/C (4.4.1); Flags (MSI|2.5G) The NIC's firmware is up to date (latest available on HP firmware update CD). ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Jonathan Feally wrote: > Attila Nagy wrote: >>>>>> Bingo, this solved the problem. The current uptime nears four days. >>>>>> Previously I couldn't go further than a day. >>>>>> >>>>>> The machine gets very light TCP load (and other machines which >>>>>> get work >>>>>> well), so I guess it's UDP RX or TX checksum related >>>>>> > I also have had my network go dead on a recent 8.0-STABLE on bge > system. Console is alive, but network just stops. I am running it as a > router with untagged on bge0 and nat of traffic on vlan201 tagged on > top of bge1. I haven't had it lock up in 3 days, but I will try the > -txcsum and -rxcsum on both interfaces to see if the problem still > persists or not. I do have a lot of tcp traffic, but there is also > unsolicited udp flying in as well. Well, it's a short time to judge from, but with rx,txcsum disabled, the machine froze nearly instantly (less than one hour of uptime), while with tso disabled, it still works. So for now I think tso causes the problems. BTW, now that we are talking about that, I remember that I've disabled it on a lot of machines previously, because I've had strange issues. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Pyun YongHyeon wrote: > On Mon, Mar 29, 2010 at 09:21:42PM +0200, Attila Nagy wrote: > >> Pyun YongHyeon wrote: >> >>> On Mon, Mar 29, 2010 at 12:57:59PM +0200, Attila Nagy wrote: >>> >>> >>>> Hi, >>>> >>>> Michael Loftis wrote: >>>> >>>> >>>>> --On Thursday, March 25, 2010 3:22 PM +0100 Attila Nagy >>>>> wrote: >>>>> >>>>> <...> >>>>> >>>>> >>>>>> Both unbound and python accepts DNS requests, and it seems when 25% >>>>>> interrupt happens, only unbound is in *udp state, where it is 50%, both >>>>>> programs are in that state. >>>>>> >>>>>> >>>>> Try turning of hardware TSO/checksum offload if it's availble on your >>>>> chipset? ifconfig -rxcsum -txcsum -tso -- I'm only using >>>>> nfe chips right now, but w/ the TSO/CSUM on they lock up constantly >>>>> under high load. We're pretty sure it's mostly the nfe driver, or the >>>>> chips themselves, but have never ruled out some generic 8.x hardware >>>>> offload issues. >>>>> >>>>> >>>> Bingo, this solved the problem. The current uptime nears four days. >>>> Previously I couldn't go further than a day. >>>> >>>> The machine gets very light TCP load (and other machines which get work >>>> well), so I guess it's UDP RX or TX checksum related. >>>> >>>> >>>> >>> Hmm, this is unexpected result. Since you're using UDP, TSO is not >>> involved in this issue. Because you disabled RX/TX checksum >>> offloading could you check how many number of 'bad checksum' and >>> and 'no checksum' you have from netstat(1)? >>> To narrow down which side of checksum offloading causes the issue, >>> would you just disable one side in a time? For instance, disable TX >>> checksum offloading with RX checksum offloading enabled and see how >>> bce(4) works. >>> #ifconfig bce0 -txcsum rxcsum >>> If that shows the same issue, try disabling RX checksum offloading >>> but enabling TX checksum offloading. >>> #ifconfig bce0 txcsum -rxcsum >>> >>> >> It's interesting. During the day, I've disabled only HW checksumming and >> left TSO enabled. It couldn't run more than a few hours. >> I have disabled tso again to see what happens. >> >> BTW, of course there is TCP traffic on that interface (DNS is also >> available on TCP), maybe this causes the problem. >> > > The only guess I can think of at this moment is incorrect use of > bus_dma(9) in TX path. But I'm not sure this is related with the > issue you're seeing. Would you try the experimental patch at the > following URL? > http://people.freebsd.org/~yongari/bce/bce.20100305.diff > Please make sure to back up your old bce(4) driver before applying > the patch. I didn't see any abnormal things in testing but it > wasn't much stressed. > With the default settings (rx, tx csum, tso) it froze in about an hour: CPU: 0.0% user, 0.0% nice, 0.0% system, 25.0% interrupt, 75.0% idle 714 bind 4 1020 1200M 1182M *lle3 17:24 0.00% unbound ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Pyun YongHyeon wrote: > On Mon, Mar 29, 2010 at 12:57:59PM +0200, Attila Nagy wrote: > >> Hi, >> >> Michael Loftis wrote: >> >>> --On Thursday, March 25, 2010 3:22 PM +0100 Attila Nagy >>> wrote: >>> >>> <...> >>> >>>> Both unbound and python accepts DNS requests, and it seems when 25% >>>> interrupt happens, only unbound is in *udp state, where it is 50%, both >>>> programs are in that state. >>>> >>> Try turning of hardware TSO/checksum offload if it's availble on your >>> chipset? ifconfig -rxcsum -txcsum -tso -- I'm only using >>> nfe chips right now, but w/ the TSO/CSUM on they lock up constantly >>> under high load. We're pretty sure it's mostly the nfe driver, or the >>> chips themselves, but have never ruled out some generic 8.x hardware >>> offload issues. >>> >> Bingo, this solved the problem. The current uptime nears four days. >> Previously I couldn't go further than a day. >> >> The machine gets very light TCP load (and other machines which get work >> well), so I guess it's UDP RX or TX checksum related. >> >> > > Hmm, this is unexpected result. Since you're using UDP, TSO is not > involved in this issue. Because you disabled RX/TX checksum > offloading could you check how many number of 'bad checksum' and > and 'no checksum' you have from netstat(1)? > To narrow down which side of checksum offloading causes the issue, > would you just disable one side in a time? For instance, disable TX > checksum offloading with RX checksum offloading enabled and see how > bce(4) works. > #ifconfig bce0 -txcsum rxcsum > If that shows the same issue, try disabling RX checksum offloading > but enabling TX checksum offloading. > #ifconfig bce0 txcsum -rxcsum > It's interesting. During the day, I've disabled only HW checksumming and left TSO enabled. It couldn't run more than a few hours. I have disabled tso again to see what happens. BTW, of course there is TCP traffic on that interface (DNS is also available on TCP), maybe this causes the problem. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Mike Tancsa wrote: > At 11:39 AM 3/25/2010, Michael Loftis wrote: >> --On Thursday, March 25, 2010 3:22 PM +0100 Attila Nagy >> wrote: >> >> <...> >>> Both unbound and python accepts DNS requests, and it seems when 25% >>> interrupt happens, only unbound is in *udp state, where it is 50%, both >>> programs are in that state. >> >> Try turning of hardware TSO/checksum offload if it's availble on your >> chipset? ifconfig -rxcsum -txcsum -tso -- I'm only using >> nfe chips right now, but w/ the TSO/CSUM on they lock up constantly >> under high load. We're pretty sure it's mostly the nfe driver, or >> the chips themselves, but have never ruled out some generic 8.x >> hardware offload issues. > > There were also a bunch of commits last night for the bce driver. If > its the NIC in RELENG_8, perhaps those bug fixes might help > > <http://lists.freebsd.org/pipermail/svn-src-stable-8/2010-March/001804.html>http://lists.freebsd.org/pipermail/svn-src-stable-8/2010-March/001804.html > > http://lists.freebsd.org/pipermail/svn-src-stable-8/2010-March/001803.html > I saw them, but they didn't seem to be related. I've just tested it, and my assumptions were correct. A fresh 8-STABLE also freezes. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Hi, Michael Loftis wrote: > > > --On Thursday, March 25, 2010 3:22 PM +0100 Attila Nagy > wrote: > > <...> >> Both unbound and python accepts DNS requests, and it seems when 25% >> interrupt happens, only unbound is in *udp state, where it is 50%, both >> programs are in that state. > > Try turning of hardware TSO/checksum offload if it's availble on your > chipset? ifconfig -rxcsum -txcsum -tso -- I'm only using > nfe chips right now, but w/ the TSO/CSUM on they lock up constantly > under high load. We're pretty sure it's mostly the nfe driver, or the > chips themselves, but have never ruled out some generic 8.x hardware > offload issues. Bingo, this solved the problem. The current uptime nears four days. Previously I couldn't go further than a day. The machine gets very light TCP load (and other machines which get work well), so I guess it's UDP RX or TX checksum related. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: 8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Pyun YongHyeon wrote: > On Thu, Mar 25, 2010 at 03:22:04PM +0100, Attila Nagy wrote: > >> Hi, >> >> I have some recursive nameservers, running unbound and 7.2-STABLE #0: >> Wed Sep 2 13:37:17 CEST 2009 on a bunch of HP BL460c machines (bce >> interfaces). >> These work OK. >> >> During the process of migrating to 8.x, I've upgraded one of these >> machines to 8.0-STABLE #25: Tue Mar 9 18:15:34 CET 2010 (the dates >> indicate an approximate time, when the source was checked out from >> cvsup.hu.freebsd.org, I don't know the exact revision). >> >> The first problem was that the machine occasionally lost network access >> for some minutes. I could log in on the console, and I could see the >> processes, involved in network IO in "keglim" state, but couldn't do any >> network IO. This lasted for some minutes, then everything came back to >> normal. >> I could fix this issue by raising kern.ipc.nmbclusters to 51200 >> (doubling from its default size), when I can't see these blackouts. >> >> But now the machine freezes. It can run for about a day, and then it >> just freezes. I can't even break in to the debugger with sending NMI to it. >> top says: >> last pid: 92428; load averages: 0.49, 0.40, 0.38up 0+21:13:18 >> 07:41:43 >> 43 processes: 2 running, 38 sleeping, 1 zombie, 2 lock >> CPU: 1.3% user, 0.0% nice, 1.3% system, 26.0% interrupt, 71.3% idle >> Mem: 1682M Active, 99M Inact, 227M Wired, 5444K Cache, 44M Buf, 5899M Free >> Swap: >> >> PID USERNAME THR PRI NICE SIZERES STATE C TIME WCPU COMMAND >> 45011 bind 4 490 1734M 1722M RUN 2 37:42 22.17% unbound >> 712 bind 3 440 70892K 19904K uwait 0 71:07 3.86% >> python2.6 >> >> The common in these freezes seems to be the high interrupt count. >> Normally, during load the CPU times look like this: >> CPU: 3.5% user, 0.0% nice, 1.8% system, 0.4% interrupt, 94.4% idle >> >> I could observe a "freeze", where top remained running and everything >> was 0%, except interrupt, which was 25% exactly (the machine has four >> cores), and another, where I could save the following console output: >> CPU: 0.0% user, 0.0% nice, 0.2% system, 50.0% interrupt, 49.8% idle >> > > When you see high number of interrupts, could you check this comes > from bce(4)? I guess you can use systat(1) to check how many number > interrupts are generated from bce(4). > I've tried it multiple times, but couldn't yet catch the moment when the machine was still alive (so the script could run) and there were increased amount of interrupts. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
8-STABLE freezes on UDP traffic (DNS), 7.x doesn't
Hi, I have some recursive nameservers, running unbound and 7.2-STABLE #0: Wed Sep 2 13:37:17 CEST 2009 on a bunch of HP BL460c machines (bce interfaces). These work OK. During the process of migrating to 8.x, I've upgraded one of these machines to 8.0-STABLE #25: Tue Mar 9 18:15:34 CET 2010 (the dates indicate an approximate time, when the source was checked out from cvsup.hu.freebsd.org, I don't know the exact revision). The first problem was that the machine occasionally lost network access for some minutes. I could log in on the console, and I could see the processes, involved in network IO in "keglim" state, but couldn't do any network IO. This lasted for some minutes, then everything came back to normal. I could fix this issue by raising kern.ipc.nmbclusters to 51200 (doubling from its default size), when I can't see these blackouts. But now the machine freezes. It can run for about a day, and then it just freezes. I can't even break in to the debugger with sending NMI to it. top says: last pid: 92428; load averages: 0.49, 0.40, 0.38up 0+21:13:18 07:41:43 43 processes: 2 running, 38 sleeping, 1 zombie, 2 lock CPU: 1.3% user, 0.0% nice, 1.3% system, 26.0% interrupt, 71.3% idle Mem: 1682M Active, 99M Inact, 227M Wired, 5444K Cache, 44M Buf, 5899M Free Swap: PID USERNAME THR PRI NICE SIZERES STATE C TIME WCPU COMMAND 45011 bind 4 490 1734M 1722M RUN 2 37:42 22.17% unbound 712 bind 3 440 70892K 19904K uwait 0 71:07 3.86% python2.6 The common in these freezes seems to be the high interrupt count. Normally, during load the CPU times look like this: CPU: 3.5% user, 0.0% nice, 1.8% system, 0.4% interrupt, 94.4% idle I could observe a "freeze", where top remained running and everything was 0%, except interrupt, which was 25% exactly (the machine has four cores), and another, where I could save the following console output: CPU: 0.0% user, 0.0% nice, 0.2% system, 50.0% interrupt, 49.8% idle ...(partial, broken line)32M 2423M *udp1 50:16 10.89% unbound 714 bind 3 440 70892K 26852K uwait 3 8:41 4.69% python2.6 61004 root 1 620 37428K 10876K *udp1 0:00 1.56% python 706 root 1 440 2696K 624K piperd 1 0:07 0.00% readproctit Both unbound and python accepts DNS requests, and it seems when 25% interrupt happens, only unbound is in *udp state, where it is 50%, both programs are in that state. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: NFS on ZFS
Hello, I've also ran into it, it's a pretty "killer" feature. :-O Any chance for us on the fix? Thanks, Kip Macy wrote: The flags checks are too strict. File a PR. I'll fix it when I get to it. Sorrry. -Kip On Wed, May 27, 2009 at 7:24 PM, Mike Andrews wrote: On Tue, 26 May 2009, Mike Andrews wrote: Takahashi Yoshihiro wrote: Today's stable has a problem creating a new file via NFS on ZFS. On the NFS server, there is no problem. % cd /ZFS % mktemp hoge hoge % ls -l hoge -rw--- 1 nyan nyan 0 5 26 19:09 hoge But it's a problem on the NFS client. # mount server:/ZFS /ZFS % cd /ZFS % mktemp hoge mktemp: mkstemp failed on hoge: Input/output error % ls -l hoge -- 1 nyan wheel 0 5 26 19:09 hoge The file has a wrong permission. This problem is only on stable, current has no problem. I'm seeing this too. It seems so far to be limited to mkstemp() -- just copying files normally works. For example /usr/bin/install -S fails, without -S works, if the target is an NFS+ZFS volume. Anyone? I've verified that if the NFS server uses UFS2, mkstemp() from an NFS client to the server works fine, but if the NFS server uses ZFS, the NFS server returns EIO after creating a file with 000 permissions. In addition to breaking /usr/bin/install -S, it also breaks rsync over NFS. I don't yet know if it matters whether the on-disk format is ZFS v6 vs v13. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
stat() takes 54 msec in a directory with 94k files (even with a big dirhash)
Hello, I have a strange error on FreeBSD 7-STABLE (compiled on 7th May, just few commits after the release, but an earlier kernel did the same). I'm doing several parallel rsyncs from a machine to another (let's call them source and destination). The source contains maildirs, so there are some directories with a (relatively) lot of files. The source runs an earlier (around 6.2) FreeBSD and plain softupdates mounted UFS2 file systems. The destination has a bigger (UFS2) filesystem, on top of gjournal, mounted as async. I've noticed that rsync sometimes stops moving data and the destination machine gets sluggish. After some testing, I could catch the effect in action (was not that hard, because it persists even for hours sometimes). top shows around 20% system activity (there are two quad core CPUs) and 0% user. The WCPU field at rsync shows 100%. ktrace-ing the rsync process I can see this: 31639 rsync0.04 CALL lstat(0x7fffab70,0x7fffaf70) 31639 rsync0.04 NAMI "hm33/00/16/uid/Maildir/new/1212536121.54673,S=3128" 31639 rsync0.054226 STRU struct stat {dev=100, ino=136943662, mode=-rw--- , nlink=1, uid=999, gid=999, rdev=546942760, atime=1241807071, stime=1212536121, ctime=1241807071, birthtime=1212536121, size=3128, blksize=4096, blocks=8, flags=0x0 } 31639 rsync0.13 RET lstat 0 31639 rsync0.18 CALL lstat(0x7fffab70,0x7fffaf70) 31639 rsync0.04 NAMI "hm33/00/16/uid/Maildir/new/1212537276.69702,S=4634" 31639 rsync0.054409 STRU struct stat {dev=100, ino=136943663, mode=-rw--- , nlink=1, uid=999, gid=999, rdev=546942762, atime=1241807071, stime=1212537276, ctime=1241807071, birthtime=1212537276, size=4634, blksize=4096, blocks=12, flags=0x0 } 31639 rsync0.13 RET lstat 0 31639 rsync0.20 CALL lstat(0x7fffab70,0x7fffaf70) 31639 rsync0.05 NAMI "hm33/00/16/uid/Maildir/new/1212537689.74390,S=3172" 31639 rsync0.054230 STRU struct stat {dev=100, ino=136943664, mode=-rw--- , nlink=1, uid=999, gid=999, rdev=546942765, atime=1241807071, stime=1212537689, ctime=1241807071, birthtime=1212537689, size=3172, blksize=4096, blocks=8, flags=0x0 } 31639 rsync0.13 RET lstat 0 So according to ktrace, the stat call takes 54 milliseconds to return for each of the files. I have tried with the default and a pretty much raised dirhash maxmem value, but I can still get these. Currently I have: vfs.ufs.dirhash_docheck: 0 vfs.ufs.dirhash_mem: 18589428 vfs.ufs.dirhash_maxmem: 209715200 vfs.ufs.dirhash_minsize: 2560 So dirhash has space to expand. The directory in question contains 94493 files. The source machine doesn't show this behaviour. top's output on the destination machine: CPU: 0.0% user, 0.0% nice, 22.7% system, 0.0% interrupt, 77.3% idle Mem: 159M Active, 3032M Inact, 599M Wired, 47M Cache, 399M Buf, 102M Free Swap: 4096M Total, 4096M Free PID USERNAME THR PRI NICE SIZERES STATE C TIME WCPU COMMAND 31639 root 1 1180 50648K 10512K CPU0 0 2:01 100.00% rsync 634 root 1 -40 2536K 628K vlruwk 1 0:20 0.00% supervise 26760 root 1 440 25940K 3316K select 1 0:10 0.00% sshd 31640 root 1 750 87512K 8324K suspfs 4 0:10 0.00% rsync 31641 root 1 750 18904K 7124K suspfs 6 0:10 0.00% rsync 31637 root 1 750 40408K 7744K suspfs 4 0:09 0.00% rsync 31636 root 1 440 20952K 6288K select 2 0:09 0.00% rsync 31638 root 1 440 104M 8912K select 3 0:09 0.00% rsync 31635 root 1 750 80344K 7812K suspfs 4 0:09 0.00% rsync 31642 root 1 440 17940K 7624K select 1 0:04 0.00% ssh 31646 root 1 450 17940K 7656K select 1 0:03 0.00% ssh All of the rsyncs use the same file system, but with different top level directories. During this, neither of the other rsyncs can run. Any ideas about what could be done to work around this? Thanks, ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: Mounting devfs over to ZFS from fstab fails
On 2008.03.28. 23:59, Vince wrote: Attila Nagy wrote: Hello, I have some jails running on ZFS, so I have to mount devfs's into them. For this purpose, I have some similar lines in /etc/fstab: devfs /pool/jail/ldap/dev devfs rw 0 0 Where /pool is a ZFS filesystem. I'm not sure if it will have any adverse effects but changing this to devfs /pool/jail/ldap/dev devfs rw,late 0 0 Will probably fix it. My guess is that the error checking correction in the latest version in -stable picked up an error that was being incorrectly ignored before. see http://www.freebsd.org/cgi/cvsweb.cgi/src/etc/rc.d/mountcritlocal.diff?only_with_tag=RELENG_7&r1=text&tr1=1.14.2.2&r2=text&tr2=1.14 which i believe is the MFC for http://www.freebsd.org/cgi/getmsg.cgi?fetch=1314016+1316331+/usr/local/www/db/text/2008/cvs-all/20080309.cvs-all Although I cant seen a commit message in cvsweb (i'm still learning that though :) Yes, late works around the problem, thanks. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Mounting devfs over to ZFS from fstab fails
Hello, I have some jails running on ZFS, so I have to mount devfs's into them. For this purpose, I have some similar lines in /etc/fstab: devfs /pool/jail/ldap/dev devfs rw 0 0 Where /pool is a ZFS filesystem. This has worked until today -when I upgraded from a previous 7-STABLE (FreeBSD 7.0-STABLE #16: Fri Mar 7 14:30:08 CET 2008) to today's STABLE- but not anymore. The boot process fails with something like WARNING: $true wasn't set, see man rc.conf (or something similar, I don't have the exact error message, but I can reproduce if it's needed), and the problem is that the rc scripts try to mount the devfs (and nullfs) stuff to the yet unmounted ZFS, so /pool/jail/ldap/dev doesn't exist. If I create them on the root filesystem, the OS boots up fine, but of course I don't have the devfs's mounted onto ZFS, they are beneath it (umount and mount -a solves the issue). There is a similar problem with nullfs's as well. AFAIK only the following has been changed in rc.d: ./dhclient ./mountcritlocal ./mountlate neither of them seems to be able to produce this kind of malfunction. Any ideas? ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: ng_fec && pseudo-device vlan
Hello, > Why don't ng_fec (Cisco FastEtherChannel netgraph module) and > `pseudo-device vlan' (802.1q trunking) work together? Don't worry. If you will ever upgrade to 5.0, not only VLANs on FEC interfaces, but FEC itself won't work. >Cisco Switch FreeBSD > >/Port 4/1--fxp0\ /vlan0 > dot1q Trunk - EtherChannel<>--fec0---< >\Port 4/2--fxp1/ \vlan1 >\vlan2 > \vlanN Nice ASCII art. You can safely s/FreeBSD/Linux/g to get this work :~-( (BTW, is it working under Linux? I don't use Linux) > I am not skilled enough to diagnose the kernel internals to discern why > this doesn't work. I am willing to test any patches that may turn up. ng_fec is in FreeBSD's tree starting from 5.0, but not yet connected to the build. I assume it won't ever... :( If anyone willing to fix it and can understand what's going on in the kernel, I will gladly provide console (serial) access to a machine which has two fxp NICs and is connected to a FEC aware switch. Any takers? --[ Free Software ISOs - http://www.fsn.hu/?f=download ]-- Attila Nagy e-mail: [EMAIL PROTECTED] Free Software Network (FSN.HU)phone @work: +361 210 1415 (194) cell.: +3630 306 6758 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: union fs
Hello, > I've been running several dataless/diskless machines with union-fs > (for the /etc stuff) with no - apparent - problems for several months > now, so i was wondering if the commnet about it being buggy still > holds? Try to run a program on a union-mounted FS which uses sendfile() to transfer data (the easiest to try is the webfsd from the ports tree). It will send garbage, collected from various places from your harddrive. ------ Attila Nagye-mail: [EMAIL PROTECTED] Budapest Polytechnic (BMF.HU) @work: +361 210 1415 (194) H-1084 Budapest, Tavaszmezo u. 15-17. cell.: +3630 306 6758 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message
Re: What's happened with kernel option ATA_ENABLE_ATAPI_DMA?
Hello, > I'm trying to upgrade to 4.3-STABLE and I've found that my kernel > doesn't want to build because its config has wrong option > ATA_ENABLE_ATAPI_DMA. Have been this option removed in 4.3? Yes. Try the hw.ata.ata_dma or hw.ata.atapi_dma sysctl instead. ------ Attila Nagye-mail: [EMAIL PROTECTED] Budapest Polytechnic (BMF.HU) @work: +361 210 1415 (194) H-1084 Budapest, Tavaszmezo u. 15-17. cell.: +3630 306 6758 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-stable" in the body of the message