Hi btrfs folks,
I am working on btrfs filesystem on how it manages the free
space. And found out btrfs maintain a ctree which manages the physical location
of the chunks and stripes of the filesystem.
Btrfs-debug-tree also gives the information on the chunk tree
I created btrfs on single device and two device.I have attached the output of
both on running btrfs-debug-tree.
For single device sum of all the length in the chunks will add upto the total
used bytes which is expected behavior.
But for two devices sum of all lengths in the chunks does not add to the total
bytes .Am I missing something .
Also I notice that for the second device the superblock location 0x10000 is not
considered as used .
I would be really grateful if you folks can answer my query.
I hav run these tests on SLES11-sp2-x86
Kernel 3.0.13.0.27-default
If you want any further clarification I would be wiiling to help you guys on
this.
Thanks,
Santosh Hosamani
________________________________
http://www.mindtree.com/email/disclaimer.html
On Tuesday, June 12, 2012 06:53:00 AM Santosh Hosamani wrote:> Kernel 3.0.13.0.27-defaultThis kernel is very old for btrfs. Can you try with at least Linux 3.4? -- R -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Santosh Hosamani posted on Tue, 12 Jun 2012 06:53:00 +0000 as excerpted:> I am working on btrfs filesystem on how it manages the free space. [...]> I hav run these tests on SLES11-sp2-x86 Kernel 3.0.13.0.27-defaultQuick mostly boilerplate intro reply. I''m just a list regular following development without getting too technical, so won''t even a full answer. 1. As you didn''t mention the wiki and based on a couple hints in your mail including the kernel version, I''m guessing you haven''t read up on btrfs there yet. https://btrfs.wiki.kernel.org/ 2. Btrfs is of course still experimental and under heavy development, so while testing as you''re doing is great, even more than with proven filesystems, if you value your data, HAVE BACKUPS. 3. It follows from the "under heavy development" part but making it explicit: for btrfs testers, a 3.0 kernel is OLD CODE with many bugs, some critical, fixed since then, and many newly implemented features in newer code. The standard STRONG recommendation is to run at least current Linus stable, thus currently 3.4, if not the 3.5 rcs, and there''s pre-Linus-integration-branches as well, if you''re brave. Additionally, for the first time last kernel cycle (btrfs wasn''t considered stable enough to bother before that), a btrfs update was submitted to the stable tree, so if you''re staying with stable, do keep current with the stable point releases, 3.3.x in that case but now of course 3.4.x. 4. Also recommended (and following from the "heavy development" bit), btrfs testers should keep current with this list in ordered to know what bugs are already being worked on and how testers might be affected. Now a brief reply that may or may not explain the one-device/multi-device difference you''re seeing, I''m not deep enough into the technical side to know for sure. As you''ll be aware if you''ve read the wiki, btrfs defaults to single data, duplicate metadata. On a single device, the two separate metadata copies are of course stored on the same device, but where btrfs has at least two devices to work with, it tries to keep the two copies on separate devices. It may be that what you''re seeing is the lower-level implications of that difference. Of course it''s also possible that you''ve found a bug, but testing with a current kernel would help in terms of knowing whether it''s still an issue or not. As I said, 3.0 is OLD for btrfs testers. (There is however someone reporting a huge metadata imbalance with multi-device btrfs filesystems and current code, IDR whether it''s 3.4 or 3.5-rcs. Only a few gigs, single digits, of data, but something like 85 gigs of metadata, after adding a second device! Definitely looks like a bug there! That''s where keeping current with the list comes in, knowing about that sort of current issue report with current code.) -- Duncan - List replies preferred. No HTML msgs. "Every nonfree program has a lord, a master -- and if you use the program, he is your master." Richard Stallman -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 06/12/2012 08:53 AM, Santosh Hosamani wrote:> > Hi btrfs folks, > I am working on btrfs filesystem on how it manages the free space. And found out btrfs maintain a ctree which manages the physical location of the chunks and stripes of the filesystem. > Btrfs-debug-tree also gives the information on the chunk tree > > I created btrfs on single device and two device.I have attached the output of both on running btrfs-debug-tree. > For single device sum of all the length in the chunks will add upto the total used bytes which is expected behavior. > > But for two devices sum of all lengths in the chunks does not add to the total bytes .Am I missing something . > Also I notice that for the second device the superblock location 0x10000 is not considered as used . > > I would be really grateful if you folks can answer my query.It''s definitely not a bug in debug-tree, but just a problem in interpreting the result. Could you please paste the output of btrfs-debug-tree -d? This way it is easier to see what''s bothering you :)> > I hav run these tests on SLES11-sp2-x86 > Kernel 3.0.13.0.27-default > > If you want any further clarification I would be wiiling to help you guys on this. > > Thanks, > Santosh Hosamani > > > > ________________________________ > > http://www.mindtree.com/email/disclaimer.html-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Jun 12, 2012 at 06:53:00AM +0000, Santosh Hosamani wrote:> > Hi btrfs folks, > I am working on btrfs filesystem on how it manages the free space. And found out btrfs maintain a ctree which manages the physical location of the chunks and stripes of the filesystem. > Btrfs-debug-tree also gives the information on the chunk tree > > I created btrfs on single device and two device.I have attached the output of both on running btrfs-debug-tree. > For single device sum of all the length in the chunks will add upto the total used bytes which is expected behavior. > > But for two devices sum of all lengths in the chunks does not add to the total bytes .Am I missing something .Without actually seeing the details of your technique and expectations, I shall make a guess that you''re not accounting for the double-counting of RAID-1 metadata. In other words, you will find that all of the metadata device extents (or chunks) will appear twice -- once on each device. Actually, this isn''t quite right either -- what you really need to do is look at the RAID-1, RAID-10 and DUP bits in the chunk flags, add up all of those chunks, divide by two, and then add in the remaining (RAID-0 and single) chunks. That total should then add up to the total value of allocated space that you get from the output of "btrfs fi df".> Also I notice that for the second device the superblock location 0x10000 is not considered as used . > > I would be really grateful if you folks can answer my query. > > I hav run these tests on SLES11-sp2-x86 > Kernel 3.0.13.0.27-defaultThis is pretty old, but shouldn''t affect the results. It will cause reliability problems if you try running it seriously. Hugo. -- === Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk == PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk --- "There''s a Martian war machine outside -- they want to talk --- to you about a cure for the common cold."
On Tue, Jun 12, 2012 at 06:53:00AM +0000, Santosh Hosamani wrote:> > Hi btrfs folks, > I am working on btrfs filesystem on how it manages the free space. And found out btrfs maintain a ctree which manages the physical location of the chunks and stripes of the filesystem. > Btrfs-debug-tree also gives the information on the chunk tree > > I created btrfs on single device and two device.I have attached the output of both on running btrfs-debug-tree. > For single device sum of all the length in the chunks will add upto the total used bytes which is expected behavior. > > But for two devices sum of all lengths in the chunks does not add to the total bytes .Am I missing something . > Also I notice that for the second device the superblock location 0x10000 is not considered as used . > > I would be really grateful if you folks can answer my query. > > I hav run these tests on SLES11-sp2-x86 > Kernel 3.0.13.0.27-defaultYour observation about non-excluded superblock is correct and this is a known bug fixed in following maintenance update of SLES. If you''re using a distribution kernel, please file a bug at the distro bug tracker (http://bugzilla.novell.com). The SLES kernel receives regular backports of btrfs fixes and features from upstream, but keeps the version at 3.0.x which confuses the upstream community ready to point you to the latest kernel released. thanks, david -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Santosh Hosamani
2012-Jun-13 04:39 UTC
RE: Bug in btrfs-debug-tree for two or more devices.
-----Original Message----- From: Arne Jansen [mailto:sensille@gmx.net] Sent: Tuesday, June 12, 2012 10:20 PM To: Santosh Hosamani Cc: linux-btrfs@vger.kernel.org Subject: Re: Bug in btrfs-debug-tree for two or more devices. On 06/12/2012 08:53 AM, Santosh Hosamani wrote:> > Hi btrfs folks, > I am working on btrfs filesystem on how it manages the free space. And found out btrfs maintain a ctree which manages the physical location of the chunks and stripes of the filesystem. > Btrfs-debug-tree also gives the information on the chunk tree > > I created btrfs on single device and two device.I have attached the output of both on running btrfs-debug-tree. > For single device sum of all the length in the chunks will add upto the total used bytes which is expected behavior. > > But for two devices sum of all lengths in the chunks does not add to the total bytes .Am I missing something . > Also I notice that for the second device the superblock location 0x10000 is not considered as used . > > I would be really grateful if you folks can answer my query.It''s definitely not a bug in debug-tree, but just a problem in interpreting the result. Could you please paste the output of btrfs-debug-tree -d? This way it is easier to see what''s bothering you :) Btrfs fi show Label: none uuid: 23f86d1e-038a-4f5b-b87c-2ba78018135c Total devices 2 FS bytes used 28.00KB devid 2 size 3.00GB used 622.38MB path /dev/sdb2 devid 1 size 3.00GB used 642.38MB path /dev/sdb1 Btrfs v0.19+ Btrfs-debug-tree -d /dev/sdb2 root tree leaf 29364224 items 9 free space 2349 generation 4 owner 1 fs uuid 23f86d1e-038a-4f5b-b87c-2ba78018135c chunk uuid db672366-6801-4f83-99ef-2087a60bb394 item 0 key (EXTENT_TREE ROOT_ITEM 0) itemoff 3756 itemsize 239 root data bytenr 29368320 level 0 dirid 0 refs 1 gen 4 item 1 key (DEV_TREE ROOT_ITEM 0) itemoff 3517 itemsize 239 root data bytenr 29372416 level 0 dirid 0 refs 1 gen 4 item 2 key (FS_TREE INODE_REF 6) itemoff 3500 itemsize 17 inode ref index 0 namelen 7 name: default item 3 key (FS_TREE ROOT_ITEM 0) itemoff 3261 itemsize 239 root data bytenr 29360128 level 0 dirid 256 refs 1 gen 4 item 4 key (ROOT_TREE_DIR INODE_ITEM 0) itemoff 3101 itemsize 160 inode generation 3 size 0 block group 0 mode 40555 links 1 item 5 key (ROOT_TREE_DIR INODE_REF 6) itemoff 3089 itemsize 12 inode ref index 0 namelen 2 name: .. item 6 key (ROOT_TREE_DIR DIR_ITEM 2378154706) itemoff 3052 itemsize 37 location key (FS_TREE ROOT_ITEM 18446744073709551615) type 2 namelen 7 datalen 0 name: default item 7 key (CSUM_TREE ROOT_ITEM 0) itemoff 2813 itemsize 239 root data bytenr 29376512 level 0 dirid 0 refs 1 gen 4 item 8 key (DATA_RELOC_TREE ROOT_ITEM 0) itemoff 2574 itemsize 239 root data bytenr 29380608 level 0 dirid 256 refs 1 gen 4 chunk tree leaf 20971520 items 8 free space 3023 generation 4 owner 3 fs uuid 23f86d1e-038a-4f5b-b87c-2ba78018135c chunk uuid db672366-6801-4f83-99ef-2087a60bb394 item 0 key (DEV_ITEMS DEV_ITEM 1) itemoff 3897 itemsize 98 dev item devid 1 total_bytes 3221225472 bytes used 673579008 item 1 key (DEV_ITEMS DEV_ITEM 2) itemoff 3799 itemsize 98 dev item devid 2 total_bytes 3221225472 bytes used 652607488 item 2 key (FIRST_CHUNK_TREE CHUNK_ITEM 0) itemoff 3719 itemsize 80 chunk length 4194304 owner 2 type 2 num_stripes 1 stripe 0 devid 1 offset 0 item 3 key (FIRST_CHUNK_TREE CHUNK_ITEM 4194304) itemoff 3639 itemsize 80 chunk length 8388608 owner 2 type 4 num_stripes 1 stripe 0 devid 1 offset 4194304 item 4 key (FIRST_CHUNK_TREE CHUNK_ITEM 12582912) itemoff 3559 itemsize 80 chunk length 8388608 owner 2 type 1 num_stripes 1 stripe 0 devid 1 offset 12582912 item 5 key (FIRST_CHUNK_TREE CHUNK_ITEM 20971520) itemoff 3447 itemsize 112 chunk length 8388608 owner 2 type 18 num_stripes 2 stripe 0 devid 2 offset 1048576 stripe 1 devid 1 offset 20971520 item 6 key (FIRST_CHUNK_TREE CHUNK_ITEM 29360128) itemoff 3335 itemsize 112 chunk length 322109440 owner 2 type 20 num_stripes 2 stripe 0 devid 2 offset 9437184 stripe 1 devid 1 offset 29360128 item 7 key (FIRST_CHUNK_TREE CHUNK_ITEM 351469568) itemoff 3223 itemsize 112 chunk length 644218880 owner 2 type 9 num_stripes 2 stripe 0 devid 2 offset 331546624 stripe 1 devid 1 offset 351469568 device tree key (DEV_TREE ROOT_ITEM 0) leaf 29372416 items 9 free space 3338 generation 4 owner 4 fs uuid 23f86d1e-038a-4f5b-b87c-2ba78018135c chunk uuid db672366-6801-4f83-99ef-2087a60bb394 item 0 key (1 DEV_EXTENT 0) itemoff 3947 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 0 length 4194304 item 1 key (1 DEV_EXTENT 4194304) itemoff 3899 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 4194304 length 8388608 item 2 key (1 DEV_EXTENT 12582912) itemoff 3851 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 12582912 length 8388608 item 3 key (1 DEV_EXTENT 20971520) itemoff 3803 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 20971520 length 8388608 item 4 key (1 DEV_EXTENT 29360128) itemoff 3755 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 29360128 length 322109440 item 5 key (1 DEV_EXTENT 351469568) itemoff 3707 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 351469568 length 322109440 item 6 key (2 DEV_EXTENT 1048576) itemoff 3659 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 20971520 length 8388608 item 7 key (2 DEV_EXTENT 9437184) itemoff 3611 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 29360128 length 322109440 item 8 key (2 DEV_EXTENT 331546624) itemoff 3563 itemsize 48 dev extent chunk_tree 3 chunk objectid 256 chunk offset 351469568 length 322109440>when you add all the lengths in the chunk tree >devid 1 407502840 >devid 2 394919928 >may be the way I am counting the length is wrong. >how to find which all blocks are used (data or metadata) and which all blocks are free in the multiple devices. >Traversing the chunk tree is correct or is there any other alternative ?>Also I had one more doubt Is chunk tree present only in the first device or will there be a copy in all the devices?________________________________ http://www.mindtree.com/email/disclaimer.html -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Santosh Hosamani
2012-Jun-13 08:40 UTC
RE: Bug in btrfs-debug-tree for two or more devices.
-----Original Message----- From: linux-btrfs-owner@vger.kernel.org [mailto:linux-btrfs-owner@vger.kernel.org] On Behalf Of Hugo Mills Sent: Wednesday, June 13, 2012 1:37 AM To: Santosh Hosamani Cc: linux-btrfs@vger.kernel.org Subject: Re: Bug in btrfs-debug-tree for two or more devices. On Tue, Jun 12, 2012 at 06:53:00AM +0000, Santosh Hosamani wrote:> > Hi btrfs folks, > I am working on btrfs filesystem on how it manages the free space. And found out btrfs maintain a ctree which manages the physical location of the chunks and stripes of the filesystem. > Btrfs-debug-tree also gives the information on the chunk tree > > I created btrfs on single device and two device.I have attached the output of both on running btrfs-debug-tree. > For single device sum of all the length in the chunks will add upto the total used bytes which is expected behavior. > > But for two devices sum of all lengths in the chunks does not add to the total bytes .Am I missing something .Without actually seeing the details of your technique and expectations, I shall make a guess that you''re not accounting for the double-counting of RAID-1 metadata. In other words, you will find that all of the metadata device extents (or chunks) will appear twice -- once on each device. Actually, this isn''t quite right either -- what you really need to do is look at the RAID-1, RAID-10 and DUP bits in the chunk flags, add up all of those chunks, divide by two, and then add in the remaining (RAID-0 and single) chunks. That total should then add up to the total value of allocated space that you get from the output of "btrfs fi df".>>chunk tree leaf 20971520 items 8 free space 3023 generation 4 owner 3 fs uuid 23f86d1e-038a-4f5b-b87c-2ba78018135c chunk uuid db672366-6801-4f83-99ef-2087a60bb394 item 0 key (DEV_ITEMS DEV_ITEM 1) itemoff 3897 itemsize 98 dev item devid 1 total_bytes 3221225472 bytes used 673579008 item 1 key (DEV_ITEMS DEV_ITEM 2) itemoff 3799 itemsize 98 dev item devid 2 total_bytes 3221225472 bytes used 652607488 item 2 key (FIRST_CHUNK_TREE CHUNK_ITEM 0) itemoff 3719 itemsize 80 chunk length 4194304 owner 2 type 2 num_stripes 1 stripe 0 devid 1 offset 0 item 3 key (FIRST_CHUNK_TREE CHUNK_ITEM 4194304) itemoff 3639 itemsize 80 chunk length 8388608 owner 2 type 4 num_stripes 1 stripe 0 devid 1 offset 4194304 item 4 key (FIRST_CHUNK_TREE CHUNK_ITEM 12582912) itemoff 3559 itemsize 80 chunk length 8388608 owner 2 type 1 num_stripes 1 stripe 0 devid 1 offset 12582912 item 5 key (FIRST_CHUNK_TREE CHUNK_ITEM 20971520) itemoff 3447 itemsize 112 chunk length 8388608 owner 2 type 18 num_stripes 2 stripe 0 devid 2 offset 1048576 stripe 1 devid 1 offset 20971520 item 6 key (FIRST_CHUNK_TREE CHUNK_ITEM 29360128) itemoff 3335 itemsize 112 chunk length 322109440 owner 2 type 20 num_stripes 2 stripe 0 devid 2 offset 9437184 stripe 1 devid 1 offset 29360128 item 7 key (FIRST_CHUNK_TREE CHUNK_ITEM 351469568) itemoff 3223 itemsize 112 chunk length 644218880 owner 2 type 9 num_stripes 2 stripe 0 devid 2 offset 331546624 stripe 1 devid 1 offset 351469568 chunk tree will tell me where the physical stripes are there right .?Irrespective of the raid type ... correct me if I am wrong. If not how will you know which all blocks are occupied and which all block are free. Basically what I want to do is . get the used blocks of all the devices and create a bitmap of that and zero out all the free block. Then I should not overwrite the used blocks. I should be able to mount the filesystem without any error. How do I achieve that?> Also I notice that for the second device the superblock location 0x10000 is not considered as used . > > I would be really grateful if you folks can answer my query. > > I hav run these tests on SLES11-sp2-x86 Kernel 3.0.13.0.27-defaultThis is pretty old, but shouldn''t affect the results. It will cause reliability problems if you try running it seriously. Hugo. -- === Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk == PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk --- "There''s a Martian war machine outside -- they want to talk --- to you about a cure for the common cold." ________________________________ http://www.mindtree.com/email/disclaimer.html -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Santosh Hosamani
2012-Jun-13 08:43 UTC
RE: Bug in btrfs-debug-tree for two or more devices.
-----Original Message----- From: linux-btrfs-owner@vger.kernel.org [mailto:linux-btrfs-owner@vger.kernel.org] On Behalf Of Randy Barlow Sent: Tuesday, June 12, 2012 8:28 PM To: linux-btrfs@vger.kernel.org Subject: Re: Bug in btrfs-debug-tree for two or more devices. On Tuesday, June 12, 2012 06:53:00 AM Santosh Hosamani wrote:> Kernel 3.0.13.0.27-defaultThis kernel is very old for btrfs. Can you try with at least Linux 3.4? I have installed 3.4.2 kernel but still I am facing the same issue.May be my understanding of calculating the used block may be wrong. If someone could help me in understanding .It would be great. -- R -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ________________________________ http://www.mindtree.com/email/disclaimer.html -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html