[dm-devel] Announce: unlimited number of shared snapshots
Mikulas Patocka
mpatocka at redhat.com
Fri Nov 28 19:54:03 UTC 2008
Oh, sorry, there was a bug with big devices. I uploaded new patches at the
same location, try them. (there's another fix for bug when sector size is
smaller than chunk size in these new patches)
BTW. for good performance, make sure that the size of your origin
partition is aligned to the chunk size --- otherwise, there's a serious
inefficiency in the kernel; if you use it over a partition with odd number
of sectors, the kernel will split all IOs to 512 bytes and it'll be very
slow.
Mikulas
> Hi Mikulas
> Thanks for your job.
>
> In my vmware ESXi guest OS (linux-2.6.28-rc5), the multisnap was crashed.
>
> dd if=/dev/zero of=/dev/LD/snap bs=4096 count=1
> echo 0 `blockdev --getsize /dev/mapper/LD-ori` multisnapshot
> /dev/mapper/LD-ori /dev/mapper/LD-snap 4096|dmsetup create ms
>
> [ 298.050106] ------------[ cut here ]------------
> [ 298.050106] kernel BUG at drivers/md/dm-bufio.c:156!
> [ 298.050106] invalid opcode: 0000 [#1] SMP [ 298.050106] last sysfs file:
> /sys/block/sde/dev
> [ 298.050106] CPU 0 [ 298.050106] Modules linked in: dm_multisnapshot
> hangcheck_timer e1000 e1000e megaraid_sas megaraid_mbox megaraid_mm mptsas
> mptspi mptscsih mptctl mptbase dm_mod scsi_transport_sas scsi_transport_spi
> sd_mod
> [ 298.050106] Pid: 1759, comm: dmsetup Not tainted 2.6.28-rc5-1128 #1
> [ 298.050106] RIP: 0010:[<ffffffffa002dae4>] [<ffffffffa002dae4>]
> get_unclaimed_buffer+0xd4/0x130 [dm_mod]
> [ 298.050106] RSP: 0018:ffff8800165edb28 EFLAGS: 00010202
> [ 298.050106] RAX: 0000000000000004 RBX: ffff880016219f00 RCX:
> ffff880016219f10
> [ 298.050106] RDX: 0000000000000902 RSI: 0000000000000001 RDI:
> ffff880016219f40
> [ 298.050106] RBP: ffff88001624e000 R08: ffff88001624e000 R09:
> ffff8800173ca000
> [ 298.050106] R10: 0000000000000003 R11: 0000000000000000 R12:
> 0000000000000001
> [ 298.050106] R13: ffff88001624e000 R14: 000000000020c401 R15:
> ffff88001624e020
> [ 298.050106] FS: 00007fd02c0356f0(0000) GS:ffffffff80658800(0000)
> knlGS:0000000000000000
> [ 298.050106] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> [ 298.050106] CR2: 00007f47f4405590 CR3: 00000000160f6000 CR4:
> 00000000000006a0
> [ 298.050106] DR0: 0000000000000000 DR1: 0000000000000000 DR2:
> 0000000000000000
> [ 298.050106] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7:
> 0000000000000400
> [ 298.050106] Process dmsetup (pid: 1759, threadinfo ffff8800165ec000, task
> ffff880025174c20)
> [ 298.050106] Stack:
> [ 298.050106] 0000000000000000 ffff880016227000 ffff880016219f00
> ffffffffa002deab
> [ 298.050106] ffff88001624e860 ffff88001624e068 0000000000000286
> ffff8800165edc18
> [ 298.050106] 0000000000000000 ffff880025174c20 ffffffff8022b950
> 0000000000000000
> [ 298.050106] Call Trace:
> [ 298.050106] [<ffffffffa002deab>] ? dm_bufio_new_read+0x2ab/0x2f0 [dm_mod]
> [ 298.050106] [<ffffffff8022b950>] ? default_wake_function+0x0/0x10
> [ 298.050106] [<ffffffffa00e0918>] ? multisnap_origin_ctr+0x4f8/0xc90
> [dm_multisnapshot]
> [ 298.050106] [<ffffffffa002932b>] ? dm_table_add_target+0x18b/0x3c0
> [dm_mod]
> [ 298.050106] [<ffffffffa002b2ff>] ? table_load+0xaf/0x210 [dm_mod]
> [ 298.050106] [<ffffffff8027b28d>] ? __vmalloc_area_node+0xbd/0x130
> [ 298.050106] [<ffffffffa002b250>] ? table_load+0x0/0x210 [dm_mod]
> [ 298.050106] [<ffffffffa002c0d1>] ? dm_ctl_ioctl+0x251/0x2c0 [dm_mod]
> [ 298.050106] [<ffffffff8029784f>] ? vfs_ioctl+0x2f/0xa0
> [ 298.050106] [<ffffffff80297c00>] ? do_vfs_ioctl+0x340/0x470
> [ 298.050106] [<ffffffff80297d79>] ? sys_ioctl+0x49/0x80
> [ 298.050106] [<ffffffff8020c10b>] ? system_call_fastpath+0x16/0x1b
> [ 298.050106] Code: 54 a8 02 74 aa 45 85 e4 74 e9 f6 07 02 74 a0 b9 02 00 00
> 00 48 c7 c2 20 d9 02 a0 be 01 00 00 00 e8 42 fe 4b e0 eb 88 0f 0b eb fe <0f>
> 0b eb fe 31 db eb 84 45 85 e4 90 0f 84 60 ff ff ff b9 02 00 [ 298.050106] RIP
> [<ffffffffa002dae4>] get_unclaimed_buffer+0xd4/0x130 [dm_mod]
> [ 298.050106] RSP <ffff8800165edb28>
> [ 298.218349] ---[ end trace 9642e91f49f4b2b1 ]---
> #dmsetup ls
> LD-snap (254, 1)
> LD-ori (254, 0)
> ms (254, 2)
> #ls /dev/mapper/
> LD-ori LD-snap control
> #vgdisplay -v Finding all volume groups
> Finding volume group "LD"
> --- Volume group ---
> VG Name LD
> System ID Format lvm2
> Metadata Areas 8
> Metadata Sequence No 3
> VG Access read/write
> VG Status resizable
> MAX LV 0
> Cur LV 2
> Open LV 2
> Max PV 0
> Cur PV 4
> Act PV 4
> VG Size 53.98 GB
> PE Size 4.00 MB
> Total PE 13820
> Alloc PE / Size 13312 / 52.00 GB
> Free PE / Size 508 / 1.98 GB
> VG UUID r0hOuU-L4I0-V3Zy-hK10-TtJe-toBw-9BDpCY
> --- Logical volume ---
> LV Name /dev/LD/ori
> VG Name LD
> LV UUID LXOLcd-oPdk-xXoq-ZmK6-B4Qd-z8y6-Vf1WZ8
> LV Write Access read/write
> LV Status available
> # open 1
> LV Size 26.00 GB
> Current LE 6656
> Segments 1
> Allocation inherit
> Read ahead sectors auto
> - currently set to 256
> Block device 254:0
> --- Logical volume ---
> LV Name /dev/LD/snap
> VG Name LD
> LV UUID MlgpYW-AbNZ-7DMd-BEmc-RNiz-wmsa-vNiq1m
> LV Write Access read/write
> LV Status available
> # open 1
> LV Size 26.00 GB
> Current LE 6656
> Segments 4
> Allocation inherit
> Read ahead sectors auto
> - currently set to 256#vgdisplay -v Finding all volume groups
> Finding volume group "LD"
> --- Volume group ---
> VG Name LD
> System ID Format lvm2
> Metadata Areas 8
> Metadata Sequence No 3
> VG Access read/write
> VG Status resizable
> MAX LV 0
> Cur LV 2
> Open LV 2
> Max PV 0
> Cur PV 4
> Act PV 4
> VG Size 53.98 GB
> PE Size 4.00 MB
> Total PE 13820
> Alloc PE / Size 13312 / 52.00 GB
> Free PE / Size 508 / 1.98 GB
> VG UUID r0hOuU-L4I0-V3Zy-hK10-TtJe-toBw-9BDpCY
> --- Logical volume ---
> LV Name /dev/LD/ori
> VG Name LD
> LV UUID LXOLcd-oPdk-xXoq-ZmK6-B4Qd-z8y6-Vf1WZ8
> LV Write Access read/write
> LV Status available
> # open 1
> LV Size 26.00 GB
> Current LE 6656
> Segments 1
> Allocation inherit
> Read ahead sectors auto
> - currently set to 256
> Block device 254:0
> --- Logical volume ---
> LV Name /dev/LD/snap
> VG Name LD
> LV UUID MlgpYW-AbNZ-7DMd-BEmc-RNiz-wmsa-vNiq1m
> LV Write Access read/write
> LV Status available
> # open 1
> LV Size 26.00 GB
> Current LE 6656
> Segments 4
> Allocation inherit
> Read ahead sectors auto
> - currently set to 256
> Block device 254:1
> --- Physical volumes ---
> PV Name /dev/sda PV UUID
> lc8iKg-WrqG-XWyL-t6TW-TtPj-HQn1-pgbtku
> PV Status allocatable
> Total PE / Free PE 7679 / 508
> PV Name /dev/sdc PV UUID
> bbHAWq-adBj-ACcU-5Bp6-LM2Y-r5M6-9ncvGc
> PV Status allocatable
> Total PE / Free PE 2047 / 0
> PV Name /dev/sdd PV UUID
> 5B7gOQ-KKuk-XtHm-6FSr-11Q4-2p7y-lr2xGP
> PV Status allocatable
> Total PE / Free PE 2047 / 0
> PV Name /dev/sde PV UUID
> kXUvPh-gsVN-oQln-CiMu-XdR2-C1Xd-DgLtjb
> PV Status allocatable
> Total PE / Free PE 2047 / 0
>
> Block device 254:1
> --- Physical volumes ---
> PV Name /dev/sda PV UUID
> lc8iKg-WrqG-XWyL-t6TW-TtPj-HQn1-pgbtku
> PV Status allocatable
> Total PE / Free PE 7679 / 508
> PV Name /dev/sdc PV UUID
> bbHAWq-adBj-ACcU-5Bp6-LM2Y-r5M6-9ncvGc
> PV Status allocatable
> Total PE / Free PE 2047 / 0
> PV Name /dev/sdd PV UUID
> 5B7gOQ-KKuk-XtHm-6FSr-11Q4-2p7y-lr2xGP
> PV Status allocatable
> Total PE / Free PE 2047 / 0
> PV Name /dev/sde PV UUID
> kXUvPh-gsVN-oQln-CiMu-XdR2-C1Xd-DgLtjb
> PV Status allocatable
> Total PE / Free PE 2047 / 0
>
> What's wrong with my test?
>
> best regards
>
More information about the dm-devel
mailing list