Hi
i'd like to ask you to CC me because i'm not subscribed to the lists
i'm having some interesting troubles
i have lvm over soft RAID-0 with LV's formated with XFS and JFS
i can work with the JFS LV's,
but i can not with the XFS one's, i can not mount them ( no troubles
with XFS normal partitions)
so
i'd like to ask is this problem with XFS or with raid or lvm
and is there a way to fix it
thanks for your help
here is what i found in dmesg
#####################################################################################
id0: zone 2
raid0: checking ide/host2/bus0/target0/lun0/part8 ... contained as device 0
(2859456) is smallest!.
raid0: checking ide/host3/bus0/target0/lun0/part10 ... nope.
raid0: checking ide/host3/bus0/target1/lun0/part6 ... nope.
raid0: zone->nb_dev: 1, size: 249024
raid0: current zone offset: 2859456
raid0: done.
raid0 : md_size is 8064256 blocks.
raid0 : conf->smallest->size is 32128 blocks.
raid0 : nb_zone is 252.
raid0 : Allocating 2016 bytes for hash.
md: updating md0 RAID superblock on device
md: ide/host3/bus0/target1/lun0/part6 [events: 000002f3]<6>(write)
ide/host3/bus0/target1/lun0/part6's sb offset: 2610432
md: ide/host3/bus0/target0/lun0/part10 [events: 000002f3]<6>(write)
ide/host3/bus0/target0/lun0/part10's sb offset: 2594368
md: ide/host2/bus0/target0/lun0/part8 [events: 000002f3]<6>(write)
ide/host2/bus0/target0/lun0/part8's sb offset: 2859456
md: considering ide/host3/bus0/target1/lun0/part5 ...
md: adding ide/host3/bus0/target1/lun0/part5 ...
md: adding ide/host3/bus0/target0/lun0/part8 ...
md: adding ide/host2/bus0/target0/lun0/part7 ...
md: created md2
md: bind<ide/host2/bus0/target0/lun0/part7,1>
md: bind<ide/host3/bus0/target0/lun0/part8,2>
md: bind<ide/host3/bus0/target1/lun0/part5,3>
md: running:
<ide/host3/bus0/target1/lun0/part5><ide/host3/bus0/target0/lun0/part8><ide/host2/bus0/target0/lun0/part7>
md: ide/host3/bus0/target1/lun0/part5's event counter: 0000031a
md: ide/host3/bus0/target0/lun0/part8's event counter: 0000031a
md: ide/host2/bus0/target0/lun0/part7's event counter: 0000031a
md2: max total readahead window set to 744k
md2: 3 data-disks, max readahead per data-disk: 248k
raid0: looking at ide/host2/bus0/target0/lun0/part7
raid0: comparing ide/host2/bus0/target0/lun0/part7(4096448) with
ide/host2/bus0/target0/lun0/part7(4096448)
raid0: END
raid0: ==> UNIQUE
raid0: 1 zones
raid0: looking at ide/host3/bus0/target0/lun0/part8
raid0: comparing ide/host3/bus0/target0/lun0/part8(4096448) with
ide/host2/bus0/target0/lun0/part7(4096448)
raid0: EQUAL
raid0: looking at ide/host3/bus0/target1/lun0/part5
raid0: comparing ide/host3/bus0/target1/lun0/part5(4096448) with
ide/host2/bus0/target0/lun0/part7(4096448)
raid0: EQUAL
raid0: FINAL 1 zones
raid0: zone 0
raid0: checking ide/host2/bus0/target0/lun0/part7 ... contained as device 0
(4096448) is smallest!.
raid0: checking ide/host3/bus0/target0/lun0/part8 ... contained as device 1
raid0: checking ide/host3/bus0/target1/lun0/part5 ... contained as device 2
raid0: zone->nb_dev: 3, size: 12289344
raid0: current zone offset: 4096448
raid0: done.
raid0 : md_size is 12289344 blocks.
raid0 : conf->smallest->size is 12289344 blocks.
raid0 : nb_zone is 1.
raid0 : Allocating 8 bytes for hash.
md: updating md2 RAID superblock on device
md: ide/host3/bus0/target1/lun0/part5 [events: 0000031b]<6>(write)
ide/host3/bus0/target1/lun0/part5's sb offset: 4096448
md: ide/host3/bus0/target0/lun0/part8 [events: 0000031b]<6>(write)
ide/host3/bus0/target0/lun0/part8's sb offset: 4096448
md: ide/host2/bus0/target0/lun0/part7 [events: 0000031b]<6>(write)
ide/host2/bus0/target0/lun0/part7's sb offset: 4096448
md: considering ide/host3/bus0/target0/lun0/part12 ...
md: adding ide/host3/bus0/target0/lun0/part12 ...
md: adding ide/host3/bus0/target0/lun0/part6 ...
md: adding ide/host3/bus0/target0/lun0/part5 ...
md: adding ide/host3/bus0/target0/lun0/part1 ...
md: created md6
md: bind<ide/host3/bus0/target0/lun0/part1,1>
md6: WARNING: ide/host3/bus0/target0/lun0/part5 appears to be on the
same physical disk as ide/host3/bus0/target0/lun0/part1. True
protection against single-disk failure might be compromised.
md: bind<ide/host3/bus0/target0/lun0/part5,2>
md6: WARNING: ide/host3/bus0/target0/lun0/part6 appears to be on the
same physical disk as ide/host3/bus0/target0/lun0/part5. True
protection against single-disk failure might be compromised.
md: bind<ide/host3/bus0/target0/lun0/part6,3>
md6: WARNING: ide/host3/bus0/target0/lun0/part12 appears to be on the
same physical disk as ide/host3/bus0/target0/lun0/part6. True
protection against single-disk failure might be compromised.
md: bind<ide/host3/bus0/target0/lun0/part12,4>
md: running:
<ide/host3/bus0/target0/lun0/part12><ide/host3/bus0/target0/lun0/part6><ide/host3/bus0/target0/lun0/part5><ide/host3/bus0/target0/lun0/part1>
md: ide/host3/bus0/target0/lun0/part12's event counter: 00000391
md: ide/host3/bus0/target0/lun0/part6's event counter: 00000391
md: ide/host3/bus0/target0/lun0/part5's event counter: 00000391
md: ide/host3/bus0/target0/lun0/part1's event counter: 00000391
md6: max total readahead window set to 124k
md6: 1 data-disks, max readahead per data-disk: 124k
md: updating md6 RAID superblock on device
md: ide/host3/bus0/target0/lun0/part12 [events: 00000392]<6>(write)
ide/host3/bus0/target0/lun0/part12's sb offset: 9759360
md: ide/host3/bus0/target0/lun0/part6 [events: 00000392]<6>(write)
ide/host3/bus0/target0/lun0/part6's sb offset: 513984
md: ide/host3/bus0/target0/lun0/part5 [events: 00000392]<6>(write)
ide/host3/bus0/target0/lun0/part5's sb offset: 2088320
md: ide/host3/bus0/target0/lun0/part1 [events: 00000392]<6>(write)
ide/host3/bus0/target0/lun0/part1's sb offset: 2048192
md: considering ide/host2/bus0/target0/lun0/part11 ...
md: adding ide/host2/bus0/target0/lun0/part11 ...
md: adding ide/host2/bus0/target0/lun0/part6 ...
md: created md7
md: bind<ide/host2/bus0/target0/lun0/part6,1>
md7: WARNING: ide/host2/bus0/target0/lun0/part11 appears to be on the
same physical disk as ide/host2/bus0/target0/lun0/part6. True
protection against single-disk failure might be compromised.
md: bind<ide/host2/bus0/target0/lun0/part11,2>
md: running:
<ide/host2/bus0/target0/lun0/part11><ide/host2/bus0/target0/lun0/part6>
md: ide/host2/bus0/target0/lun0/part11's event counter: 000003bf
md: ide/host2/bus0/target0/lun0/part6's event counter: 000003bf
md7: max total readahead window set to 124k
md7: 1 data-disks, max readahead per data-disk: 124k
md: updating md7 RAID superblock on device
md: ide/host2/bus0/target0/lun0/part11 [events: 000003c0]<6>(write)
ide/host2/bus0/target0/lun0/part11's sb offset: 10514432
md: ide/host2/bus0/target0/lun0/part6 [events: 000003c0]<6>(write)
ide/host2/bus0/target0/lun0/part6's sb offset: 3421696
md: ... autorun DONE.
LVM version 1.0.1-rc4(ish)(03/10/2001)
IEEE 802.2 LLC for Linux 2.1 (c) 1996 Tim Alpaerts
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 4096 buckets, 32Kbytes
TCP: Hash tables configured (established 131072 bind 65536)
Linux IP multicast router 0.06 plus PIM-SM
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
found reiserfs format "3.6" with standard journal
Reiserfs journal params: device 22:4b, size 8192, journal first block
18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
reiserfs: checking transaction log (ide3(34,75)) for (ide3(34,75))
Using r5 hash to sort names
VFS: Mounted root (reiserfs filesystem) readonly.
Mounted devfs on /dev
Freeing unused kernel memory: 256k freed
Real Time Clock Driver v1.10e
loop: loaded (max 32 devices)
mice: PS/2 mouse device common for all mice
cdrom: open failed.
VFS: Disk change detected on device ide1(22,0)
Adding Swap: 473876k swap-space (priority 1)
Adding Swap: 457812k swap-space (priority 1)
Adding Swap: 473876k swap-space (priority 1)
XFS mounting filesystem lvm(58,2)
raid0_make_request bug: can't convert block across chunks or bigger than
16k 8323317 64
raid0_make_request bug: can't convert block across chunks or bigger than
16k 8323445 64
I/O error in filesystem ("lvm(58,2)") meta-data dev 0xc0223a02 block
0x601f7d
("xlog_bread") error 5 buf count 131072
raid0_make_request bug: can't convert block across chunks or bigger than
16k 8324829 29
I/O error in filesystem ("lvm(58,2)") meta-data dev 0xc0223a02 block
0x602565
("xlog_bread") error 5 buf count 30208
XFS: failed to find log head
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,3)
XFS: WARNING: recovery required on readonly filesystem.
XFS: write access will be enabled during mount.
Starting XFS recovery on filesystem: lvm(58,3) (dev: 58/3)
XFS: xlog_recover_process_data: bad clientid
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,1)
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610115 64
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610243 64
I/O error in filesystem ("lvm(58,1)") meta-data dev 0xc0223a01 block
0xa0218b
("xlog_bread") error 5 buf count 131072
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23611101 29
I/O error in filesystem ("lvm(58,1)") meta-data dev 0xc0223a01 block
0xa02565
("xlog_bread") error 5 buf count 30208
XFS: failed to find log head
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,2)
raid0_make_request bug: can't convert block across chunks or bigger than
16k 8323317 64
raid0_make_request bug: can't convert block across chunks or bigger than
16k 8323445 64
I/O error in filesystem ("lvm(58,2)") meta-data dev 0xc0223a02 block
0x601f7d
("xlog_bread") error 5 buf count 131072
raid0_make_request bug: can't convert block across chunks or bigger than
16k 8324829 29
I/O error in filesystem ("lvm(58,2)") meta-data dev 0xc0223a02 block
0x602565
("xlog_bread") error 5 buf count 30208
XFS: failed to find log head
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,3)
XFS: WARNING: recovery required on readonly filesystem.
XFS: write access will be enabled during mount.
Starting XFS recovery on filesystem: lvm(58,3) (dev: 58/3)
XFS: xlog_recover_process_data: bad clientid
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,1)
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610115 64
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610243 64
I/O error in filesystem ("lvm(58,1)") meta-data dev 0xc0223a01 block
0xa0218b
("xlog_bread") error 5 buf count 131072
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23611101 29
I/O error in filesystem ("lvm(58,1)") meta-data dev 0xc0223a01 block
0xa02565
("xlog_bread") error 5 buf count 30208
XFS: failed to find log head
XFS: log mount/recovery failed
XFS: log mount failed
svetljo wrote:
> Hi
> i'd like to ask you to CC me because i'm not subscribed to the lists
>
> i'm having some interesting troubles
> i have lvm over soft RAID-0 with LV's formated with XFS and JFS
> i can work with the JFS LV's,
> but i can not with the XFS one's, i can not mount them ( no troubles
> with XFS normal partitions)
>
> so
> i'd like to ask is this problem with XFS or with raid or lvm
> and is there a way to fix it
>
> thanks for your help
>
> here is what i found in dmesg
>
>
> XFS mounting filesystem lvm(58,2)
> raid0_make_request bug: can't convert block across chunks or bigger than
> 16k 8323317 64
> raid0_make_request bug: can't convert block across chunks or bigger than
> 16k 8323445 64
> I/O error in filesystem ("lvm(58,2)") meta-data dev 0xc0223a02 block
> 0x601f7d
> ("xlog_bread") error 5 buf count 131072
> raid0_make_request bug: can't convert block across chunks or bigger than
> 16k 8324829 29
> I/O error in filesystem ("lvm(58,2)") meta-data dev 0xc0223a02 block
> 0x602565
> ("xlog_bread") error 5 buf count 30208
>
This is your problem, in the 2.5 code base, the bio infrastructure and the
raid code do not work well together. It is being worked on - slowly.
If you want to dumb down xfs to make it function then I suspect you
can do it by editing
fs/xfs/pagebuf/page_buf.c
looking for the line which uses BIO_MAX_SECTORS and replace
nr_pages = BIO_MAX_SECTORS >> (PAGE_SHIFT - 9);
with
nr_pages = 1;
And for extra bonus points, only do it when pb->pb_dev is on the
MD_MAJOR device.
This will make xfs send smaller bio structures down to the block
layer and hopefully avoid the problem.
I have not tested this - don't have any time right now, on a plane
in 6 hours and way too much to do.
Steve
Dave Jones wrote:
>On Fri, Apr 05, 2002 at 01:33:49PM +0200, svetljo wrote:
> > i'm having some interesting troubles
> > i have lvm over soft RAID-0 with LV's formated with XFS and JFS
> > i can work with the JFS LV's,
> > but i can not with the XFS one's, i can not mount them ( no troubles
> > with XFS normal partitions)
> > so i'd like to ask is this problem with XFS or with raid or lvm
> > and is there a way to fix it
>
>IIRC, this was reported a while ago, and it was something to do with
>XFS creating too large requests that upset the raid code.
>
Or the raid code not handling the bio layer too well, depends on your point
of view ;-)
Steve
On Fri, Apr 05, 2002 at 01:33:49PM +0200, svetljo wrote:
> i'm having some interesting troubles
> i have lvm over soft RAID-0 with LV's formated with XFS and JFS
> i can work with the JFS LV's,
> but i can not with the XFS one's, i can not mount them ( no troubles
> with XFS normal partitions)
> so i'd like to ask is this problem with XFS or with raid or lvm
> and is there a way to fix it
IIRC, this was reported a while ago, and it was something to do with
XFS creating too large requests that upset the raid code.
--
| Dave Jones. http://www.codemonkey.org.uk
| SuSE Labs
Stephen Lord wrote:
>
> Dave Jones wrote:
>
> >On Fri, Apr 05, 2002 at 01:33:49PM +0200, svetljo wrote:
> > > i'm having some interesting troubles
> > > i have lvm over soft RAID-0 with LV's formated with XFS and JFS
> > > i can work with the JFS LV's,
> > > but i can not with the XFS one's, i can not mount them ( no troubles
> > > with XFS normal partitions)
> > > so i'd like to ask is this problem with XFS or with raid or lvm
> > > and is there a way to fix it
> >
> >IIRC, this was reported a while ago, and it was something to do with
> >XFS creating too large requests that upset the raid code.
> >
> Or the raid code not handling the bio layer too well, depends on your point
> of view ;-)
>
Stephen's point of view is correct. RAID0 fails
in the same manner with the large pagecache BIOs
which I'm feeding it. It fails in the same manner
with O_DIRECT on ext2.
Neil knows about it, and will get to it. mkp has
a 2.4 request splitter which he will turn into a
2.5 BIO splitter.
As you said - it's being worked on.
-
Hi
>
> This is your problem, in the 2.5 code base, the bio infrastructure
and the
> raid code do not work well together. It is being worked on - slowly.
>
> If you want to dumb down xfs to make it function then I suspect you
> can do it by editing
>
> fs/xfs/pagebuf/page_buf.c
>
> looking for the line which uses BIO_MAX_SECTORS and replace
>
> nr_pages = BIO_MAX_SECTORS >> (PAGE_SHIFT - 9);
>
> with
>
> nr_pages = 1;
>
> And for extra bonus points, only do it when pb->pb_dev is on the
> MD_MAJOR device.
>
> This will make xfs send smaller bio structures down to the block
> layer and hopefully avoid the problem.
>
> I have not tested this - don't have any time right now, on a plane
> in 6 hours and way too much to do.
>
> Steve
>
that helped a lot but i still have two LV's, witch i can not mount
the one complains again about raid0_make_request [lvm(58,1)], but the
other does not [lvm(58,5)]
XFS mounting filesystem lvm(58,5)
Starting XFS recovery on filesystem: lvm(58,5) (dev: 58/5)
XFS: xlog_recover_process_data: bad clientid
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,6)
XFS mounting filesystem lvm(58,2)
XFS mounting filesystem lvm(58,3)
XFS mounting filesystem lvm(58,1)
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610013 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610045 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610077 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610109 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610141 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610173 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610205 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23610237 4
I/O error in filesystem ("lvm(58,1)") meta-data dev 0xc03d3a01 block
0xa0210d
("xlog_bread") error 5 buf count 131072
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23611101 4
raid0_make_request bug: can't convert block across chunks or bigger than
16k 23611133 4
I/O error in filesystem ("lvm(58,1)") meta-data dev 0xc03d3a01 block
0xa02565
("xlog_bread") error 5 buf count 30208
XFS: failed to find log head
XFS: log mount/recovery failed
XFS: log mount failed
XFS mounting filesystem lvm(58,4)