| Message ID | cover.1735748715.git.anand.jain@oracle.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | raid1 balancing methods | expand |
On Thu, Jan 02, 2025 at 02:06:29AM +0800, Anand Jain wrote: > v5: > Fixes based on review comments: > . Rewrite `btrfs_read_policy_to_enum()` using `sysfs_match_string()` > and `strncpy()`. > . Rewrite the round-robin method based on read counts. > . Fix the smatch indentation warning. > - Change the default minimum contiguous device read size for round-robin > from 256K to 192K, as the latter performs slightly better. This depends on the device and load and the read pattern so any number is fine, I'd rather stick to something that looks sensible if the difference is slight. Changing that later based on extensive benchmarks. > - Introduce a framework to track filesystem read counts. (New patch) > - Reran defrag performance numbers > $ xfs_io -f -d -c 'pwrite -S 0xab 0 1000000000' /btrfs/P6B > $ time -p btrfs filesystem defrag -r -f -c /btrfs > > | | Time | Read I/O Count | gain | > | | Real | devid1 | devid2 | w-PID | > |---------|-------|--------|--------|-------| > | pid | 11.14s| 3808 | 0 | - | > | rotation| | | | | > | 196608| 6.54s| 2532 | 1276 | 41.29%| > | 262144| 8.42s| 1907 | 1901 | 24.41%| > | devid:1 | 10.95s| 3807 | 0 | 1.70% | > > v4: As we're at rc5 we need to get this series to for-next soon, before rc6 is released this Sunday. I'll do one review pass and then move the patches to for-next. I'm not convinced we should do the module parameter but as this is to ease testing let's do it.
On 2/1/25 19:28, David Sterba wrote: > On Thu, Jan 02, 2025 at 02:06:29AM +0800, Anand Jain wrote: >> v5: >> Fixes based on review comments: >> . Rewrite `btrfs_read_policy_to_enum()` using `sysfs_match_string()` >> and `strncpy()`. >> . Rewrite the round-robin method based on read counts. >> . Fix the smatch indentation warning. >> - Change the default minimum contiguous device read size for round-robin >> from 256K to 192K, as the latter performs slightly better. > > This depends on the device and load and the read pattern so any number > is fine, I'd rather stick to something that looks sensible if the > difference is slight. Changing that later based on extensive benchmarks. > The optimization should target the generic read-write workload, where I find 256k or more reasonable, though I am unable to demonstrate this at the moment. It's a good idea to gather more feedback. I've sent a fix with SZ_256K and for a missed bug, hoping both are folded into their patches. Thx. >> - Introduce a framework to track filesystem read counts. (New patch) >> - Reran defrag performance numbers >> $ xfs_io -f -d -c 'pwrite -S 0xab 0 1000000000' /btrfs/P6B >> $ time -p btrfs filesystem defrag -r -f -c /btrfs >> >> | | Time | Read I/O Count | gain | >> | | Real | devid1 | devid2 | w-PID | >> |---------|-------|--------|--------|-------| >> | pid | 11.14s| 3808 | 0 | - | >> | rotation| | | | | >> | 196608| 6.54s| 2532 | 1276 | 41.29%| >> | 262144| 8.42s| 1907 | 1901 | 24.41%| >> | devid:1 | 10.95s| 3807 | 0 | 1.70% | >> >> v4: > > As we're at rc5 we need to get this series to for-next soon, before rc6 > is released this Sunday. I'll do one review pass and then move the > patches to for-next. I'm not convinced we should do the module parameter > but as this is to ease testing let's do it. Thanks, Anand
v5: Fixes based on review comments: . Rewrite `btrfs_read_policy_to_enum()` using `sysfs_match_string()` and `strncpy()`. . Rewrite the round-robin method based on read counts. . Fix the smatch indentation warning. - Change the default minimum contiguous device read size for round-robin from 256K to 192K, as the latter performs slightly better. - Introduce a framework to track filesystem read counts. (New patch) - Reran defrag performance numbers $ xfs_io -f -d -c 'pwrite -S 0xab 0 1000000000' /btrfs/P6B $ time -p btrfs filesystem defrag -r -f -c /btrfs | | Time | Read I/O Count | gain | | | Real | devid1 | devid2 | w-PID | |---------|-------|--------|--------|-------| | pid | 11.14s| 3808 | 0 | - | | rotation| | | | | | 196608| 6.54s| 2532 | 1276 | 41.29%| | 262144| 8.42s| 1907 | 1901 | 24.41%| | devid:1 | 10.95s| 3807 | 0 | 1.70% | v4: Fixes based on review comments: 3/10: Use == 0 to check strlen(str); drop dynamic alloc for %param. 4/10: Add __maybe_unused for %value_str in btrfs_read_policy_to_enum(). Return int instead of enum btrfs_read_policy. 5/10: Fix change-log (: is part of optional [ ]). Wrap btrfs_read_policy_name[] with ifdef for new methods. Use IS_ALIGNED() for sector-size alignment check. Roundup misaligned %value. Use named constants: BTRFS_DEFAULT_RR_MIN_CONTIGUOUS_READ, BTRFS_RAID1_MAX_MIRRORS. Mark %s1 and %s2 in btrfs_cmp_devid() as const. Add comments to btrfs_read_rr(); Use loop-local %i. Add space around /. Use >> for sector-size division. Prefix %min_contiguous_read with rr. 7/10: Move experimental to the top of the feature list. Use experiment=on. Skip printing when off. v3: 1. Removed the latency-based RAID1 balancing patch. (Per David's review) 2. Renamed "rotation" to "round-robin" and set the per-set min_contiguous_read to 256k. (Per David's review) 3. Added raid1-balancing module configuration for fstests testing. raid1-balancing can now be configured through both module load parameters and sysfs. The logic of individual methods remains unchanged, and performance metrics are consistent with v2. ----- v2: 1. Move new features to CONFIG_BTRFS_EXPERIMENTAL instead of CONFIG_BTRFS_DEBUG. 2. Correct the typo from %est_wait to %best_wait. 3. Initialize %best_wait to U64_MAX and remove the check for 0. 4. Implement rotation with a minimum contiguous read threshold before switching to the next stripe. Configure this, using: echo rotation:[min_contiguous_read] > /sys/fs/btrfs/<uuid>/read_policy The default value is the sector size, and the min_contiguous_read value must be a multiple of the sector size. 5. Tested FIO random read/write and defrag compression workloads with min_contiguous_read set to sector size, 192k, and 256k. RAID1 balancing method rotation is better for multi-process workloads such as fio and also single-process workload such as defragmentation. $ fio --filename=/btrfs/foo --size=5Gi --direct=1 --rw=randrw --bs=4k \ --ioengine=libaio --iodepth=256 --runtime=120 --numjobs=4 \ --time_based --group_reporting --name=iops-test-job --eta-newline=1 | | | | Read I/O count | | | Read | Write | devid1 | devid2 | |---------|------------|------------|--------|--------| | pid | 20.3MiB/s | 20.5MiB/s | 313895 | 313895 | | rotation| | | | | | 4096| 20.4MiB/s | 20.5MiB/s | 313895 | 313895 | | 196608| 20.2MiB/s | 20.2MiB/s | 310152 | 310175 | | 262144| 20.3MiB/s | 20.4MiB/s | 312180 | 312191 | | latency| 18.4MiB/s | 18.4MiB/s | 272980 | 291683 | | devid:1 | 14.8MiB/s | 14.9MiB/s | 456376 | 0 | rotation RAID1 balancing technique performs more than 2x better for single-process defrag. $ time -p btrfs filesystem defrag -r -f -c /btrfs | | Time | Read I/O Count | | | Real | devid1 | devid2 | |---------|-------|--------|--------| | pid | 18.00s| 3800 | 0 | | rotation| | | | | 4096| 8.95s| 1900 | 1901 | | 196608| 8.50s| 1881 | 1919 | | 262144| 8.80s| 1881 | 1919 | | latency | 17.18s| 3800 | 0 | | devid:2 | 17.48s| 0 | 3800 | Rotation keeps all devices active, and for now, the Rotation RAID1 balancing method is preferable as default. More workload testing is needed while the code is EXPERIMENTAL. While Latency is better during the failing/unstable block layer transport. As of now these two techniques, are needed to be further independently tested with different worloads, and in the long term we should be merge these technique to a unified heuristic. Rotation keeps all devices active, and for now, the Rotation RAID1 balancing method should be the default. More workload testing is needed while the code is EXPERIMENTAL. Latency is smarter with unstable block layer transport. Both techniques need independent testing across workloads, with the goal of eventually merging them into a unified approach? for the long term. Devid is a hands-on approach, provides manual or user-space script control. These RAID1 balancing methods are tunable via the sysfs knob. The mount -o option and btrfs properties are under consideration. Thx. --------- original v1 ------------ The RAID1-balancing methods helps distribute read I/O across devices, and this patch introduces three balancing methods: rotation, latency, and devid. These methods are enabled under the `CONFIG_BTRFS_DEBUG` config option and are on top of the previously added `/sys/fs/btrfs/<UUID>/read_policy` interface to configure the desired RAID1 read balancing method. I've tested these patches using fio and filesystem defragmentation workloads on a two-device RAID1 setup (with both data and metadata mirrored across identical devices). I tracked device read counts by extracting stats from `/sys/devices/<..>/stat` for each device. Below is a summary of the results, with each result the average of three iterations. A typical generic random rw workload: $ fio --filename=/btrfs/foo --size=10Gi --direct=1 --rw=randrw --bs=4k \ --ioengine=libaio --iodepth=256 --runtime=120 --numjobs=4 --time_based \ --group_reporting --name=iops-test-job --eta-newline=1 | | | | Read I/O count | | | Read | Write | devid1 | devid2 | |---------|------------|------------|--------|--------| | pid | 29.4MiB/s | 29.5MiB/s | 456548 | 447975 | | rotation| 29.3MiB/s | 29.3MiB/s | 450105 | 450055 | | latency | 21.9MiB/s | 21.9MiB/s | 672387 | 0 | | devid:1 | 22.0MiB/s | 22.0MiB/s | 674788 | 0 | Defragmentation with compression workload: $ xfs_io -f -d -c 'pwrite -S 0xab 0 1G' /btrfs/foo $ sync $ echo 3 > /proc/sys/vm/drop_caches $ btrfs filesystem defrag -f -c /btrfs/foo | | Time | Read I/O Count | | | Real | devid1 | devid2 | |---------|-------|--------|--------| | pid | 21.61s| 3810 | 0 | | rotation| 11.55s| 1905 | 1905 | | latency | 20.99s| 0 | 3810 | | devid:2 | 21.41s| 0 | 3810 | . The PID-based balancing method works well for the generic random rw fio workload. . The rotation method is ideal when you want to keep both devices active, and it boosts performance in sequential defragmentation scenarios. . The latency-based method work well when we have mixed device types or when one device experiences intermittent I/O failures the latency increases and it automatically picks the other device for further Read IOs. . The devid method is a more hands-on approach, useful for diagnosing and testing RAID1 mirror synchronizations. Anand Jain (10): btrfs: initialize fs_devices->fs_info earlier btrfs: simplify output formatting in btrfs_read_policy_show btrfs: add btrfs_read_policy_to_enum helper and refactor read policy store btrfs: handle value associated with raid1 balancing parameter btrfs: add read count tracking for filesystem stats btrfs: introduce RAID1 round-robin read balancing btrfs: add RAID1 preferred read device btrfs: expose experimental mode in module information btrfs: enable RAID1 balancing configuration via modprobe parameter btrfs: modload to print RAID1 balancing status fs/btrfs/bio.c | 8 +++ fs/btrfs/disk-io.c | 4 ++ fs/btrfs/super.c | 18 +++++ fs/btrfs/sysfs.c | 168 ++++++++++++++++++++++++++++++++++++++++----- fs/btrfs/sysfs.h | 5 ++ fs/btrfs/volumes.c | 115 ++++++++++++++++++++++++++++++- fs/btrfs/volumes.h | 22 +++++- 7 files changed, 317 insertions(+), 23 deletions(-)