Nowadays, I don't know about correctness of my RAID1. Misunderstanding with device usage. Confused by different DATA size in RAID1 devices. btrfs balance start -dconvert=raid1 -mconvert=raid1 /mnt changed device usage, but not realign values. I unmounted and checked devices by check command with error = 0 message. btrfs fi df -kh . Data, RAID1: total=847.00GiB, used=842.37GiB System, RAID1: total=32.00MiB, used=160.00KiB Metadata, RAID1: total=4.00GiB, used=2.48GiB GlobalReserve, single: total=512.00MiB, used=0.00B btrfs fi show Label: 'exchange' uuid: 052d18f5-8996-4935-a474-30886714a980 Total devices 1 FS bytes used 288.72GiB devid 1 size 922.19GiB used 291.04GiB path /dev/sdc2 Label: 'storage' uuid: b8dddbed-911d-457f-9101-897a13c215bd Total devices 3 FS bytes used 844.85GiB >>> devid 1 size 1.82TiB used 564.03GiB path /dev/sdb >>> devid 2 size 1.82TiB used 564.00GiB path /dev/sde >>> devid 3 size 1.83TiB used 574.03GiB path /dev/sdc1 btrfs device usage . /dev/sdb, ID: 1 Device size: 1.82TiB >>> Data,RAID1: 561.00GiB Metadata,RAID1: 3.00GiB System,RAID1: 32.00MiB Unallocated: 1.27TiB /dev/sdc1, ID: 3 Device size: 1.83TiB >>> Data,RAID1: 571.00GiB Metadata,RAID1: 3.00GiB System,RAID1: 32.00MiB Unallocated: 1.27TiB /dev/sde, ID: 2 Device size: 1.82TiB >>> Data,RAID1: 562.00GiB Metadata,RAID1: 2.00GiB Unallocated: 1.27TiB blkid | grep btrfs /dev/sdb: LABEL="storage" UUID="b8dddbed-911d-457f-9101-897a13c215bd" UUID_SUB="4e2e90ad-d18c-4e89-93ff-6fcbc85e3aae" TYPE="btrfs" /dev/sdc1: LABEL="storage" UUID="b8dddbed-911d-457f-9101-897a13c215bd" UUID_SUB="6fcd7643-d586-4c25-862b-0fde7f278536" TYPE="btrfs" PARTUUID="40bbcc22-5c44-417e-8daf-4537cb4ba2df" /dev/sdc2: LABEL="exchange" UUID="052d18f5-8996-4935-a474-30886714a980" UUID_SUB="7052acc4-ef49-46b0-b082-0dbb5103e34f" TYPE="btrfs" PARTUUID="36714c06-ca8f-4668-8d02-60e7e24b136a" /dev/sde: LABEL="storage" UUID="b8dddbed-911d-457f-9101-897a13c215bd" UUID_SUB="76c5d5b6-4c5d-4b89-b58f-85e876c465ec" TYPE="btrfs" mount | grep /d/d /dev/sdb on /d/d type btrfs (rw,relatime,compress-force=zlib,space_cache,autodefrag)
btrfs fi usage . Overall: Device size: 5.47TiB Device allocated: 1.66TiB Device unallocated: 3.80TiB Device missing: 0.00B Used: 1.65TiB Free (estimated): 1.91TiB (min: 1.91TiB) Data ratio: 2.00 Metadata ratio: 2.00 Global reserve: 512.00MiB (used: 0.00B) Data,RAID1: Size:847.00GiB, Used:842.37GiB /dev/sdb 561.00GiB /dev/sdc1 571.00GiB /dev/sde 562.00GiB Metadata,RAID1: Size:4.00GiB, Used:2.48GiB /dev/sdb 3.00GiB /dev/sdc1 3.00GiB /dev/sde 2.00GiB System,RAID1: Size:32.00MiB, Used:160.00KiB /dev/sdb 32.00MiB /dev/sdc1 32.00MiB Unallocated: /dev/sdb 1.27TiB /dev/sdc1 1.27TiB /dev/sde 1.27TiB
Devices sdb and sde is the same model of hard disk smartctl -a /dev/sdb >f ; smartctl -a /dev/sdb >s ; diff f s shows no differences === START OF INFORMATION SECTION === Model Family: Western Digital AV-GP (AF) Device Model: WDC WD20EURS-63S48Y0 Serial Number: WD-WMAZA7578083 LU WWN Device Id: 5 0014ee 2b16830ee Firmware Version: 51.0AB51 User Capacity: 2 000 398 934 016 bytes [2,00 TB] Sector Sizes: 512 bytes logical, 4096 bytes physical Device is: In smartctl database [for details use: -P show] ATA Version is: ATA8-ACS (minor revision not indicated) SATA Version is: SATA 2.6, 3.0 Gb/s Local Time is: Sun May 17 21:39:46 2015 MSK SMART support is: Available - device has SMART capability. SMART support is: Enabled
btrfs balance start -v -dusage=99 /d/d and btrfs balance start -v /d/d gave me /dev/sdb, ID: 1 Device size: 1.82TiB Data,RAID1: 560.00GiB Metadata,RAID1: 2.00GiB Unallocated: 1.27TiB /dev/sdc1, ID: 3 Device size: 1.83TiB Data,RAID1: 567.00GiB Metadata,RAID1: 3.00GiB System,RAID1: 32.00MiB Unallocated: 1.27TiB /dev/sde, ID: 2 Device size: 1.82TiB Data,RAID1: 557.00GiB Metadata,RAID1: 3.00GiB System,RAID1: 32.00MiB Unallocated: 1.27TiB so, metadata and data is different across three raid1 volumes again...
If this RAID0 with RAID1 label? o_O Please, review this bug ASAP, or I'll reformat drives...
The numbers look ok with respect to what raid1 means on btrfs: there always 2 copies of the given data, not N like for md-raid1 does. Therefore: * from 'btrfs fi df' output Metadata, RAID1: total=4.00GiB, used=2.48GiB there are 4G of total space (in logical units), that has to be saved twice, which gives 8G * from 'btrfs fi usage' output, comment 3: sdb: 2G sdc1: 3G sde: 3G sums to 8G as expected.
Sorry, I don't know about this behaviour of btrfs... All my collegues don't knows too... Sorry for question, but the only way to get 3 copies is to remove one drive at sync it by rsync? Or any property of btrfs is able to do this?
Sorry, resolved by one deleted drive.