Sparse files contain large regions containing only zero bytes. Checking
that a blob only contains zeros is possible with over 100GB/s for modern
x86 CPUs. Calculating sha256 hashes is only possible with 500MB/s (or
2GB/s using hardware acceleration). Thus we can speed up the hash
calculation for all zero blobs (which always have length
chunker.MinSize) by checking for zero bytes and then using the
precomputed hash.
The all zeros check is only performed for blobs with the minimal chunk
size, and thus should add no overhead most of the time. For chunks which
are not all zero but have the minimal chunks size, the overhead will be
below 2% based on the above performance numbers.
This allows reading sparse sections of files as fast as the kernel can
return data to us. On my system using BTRFS this resulted in about
4GB/s.
The restorer can issue multiple calls to WriteAt in parallel. This can
result in unexpected orderings of the Truncate and WriteAt calls and
sometimes too short restored files.
We can either preallocate storage for a file or sparsify it. This
detects a pack file as sparse if it contains an all zero block or
consists of only one block. As the file sparsification is just an
approximation, hide it behind a `--sparse` parameter.
This writes files by using (*os.File).Truncate, which resolves to the
truncate system call on Unix.
Compared to the naive loop,
for _, b := range p {
if b != 0 {
return false
}
}
the optimized allZero is about 10× faster:
name old time/op new time/op delta
AllZero-8 1.09ms ± 1% 0.09ms ± 1% -92.10% (p=0.000 n=10+10)
name old speed new speed delta
AllZero-8 3.84GB/s ± 1% 48.59GB/s ± 1% +1166.51% (p=0.000 n=10+10)
Use runtime.GOMAXPROCS(0) as worker count for CPU-bound tasks,
repo.Connections() for IO-bound task and a combination if a task can be
both. Streaming packs is treated as IO-bound as adding more worker
cannot provide a speedup.
Typical IO-bound tasks are download / uploading / deleting files.
Decoding / Encoding / Verifying are usually CPU-bound. Several tasks are
a combination of both, e.g. for combined download and decode functions.
In the latter case add both limits together. As the backends have their
own concurrency limits restic still won't download more than
repo.Connections() files in parallel, but the additional workers can
decode already downloaded data in parallel.
Previously, SaveAndEncrypt would assemble blobs into packs and either
return immediately if the pack is not yet full or upload the pack file
otherwise. The upload will block the current goroutine until it
finishes.
Now, the upload is done using separate goroutines. This requires changes
to the error handling. As uploads are no longer tied to a SaveAndEncrypt
call, failed uploads are signaled using an errgroup.
To count the uploaded amount of data, the pack header overhead is no
longer returned by `packer.Finalize` but rather by
`packer.HeaderOverhead`. This helper method is necessary to continue
returning the pack header overhead directly to the responsible call to
`repository.SaveBlob`. Without the method this would not be possible,
as packs are finalized asynchronously.
Failed pack/blob downloads should be retried. For blobs that fail
decryption assume that the pack file is really damaged and try to
restore the remaining blobs.
Michael Eischer
greatroar
Alexander Neumann
Alexander Neumann
Alexander Weiss