Re: [Libguestfs] [PATCH libnbd] PROPOSAL Add nbdcp (NBD copying) tool.

This is a proposal for an NBD to/from file copying tool (not actually written). Obviously this would duplicate functionality which is already available in qemu-img convert. The reasons for writing this tool would be: - to produce a tool which is very focused on the specific needs of virt-v2v and similar migration scenarios - to have a small tool with minimal dependencies - fix some of the obvious problems with qemu-img convert like how it handles devices which are already zeroed, and preallocation - enable optimizations which make sense for NBD but which might not be applicable to qemu (eg. multi-conn, freely writing out of order from multiple threads). Alternative is of course to not write a new tool and instead try to fix all this stuff in qemu-img itself. Anyway let me know your thoughts.

Rich. nbdcp(1) LIBNBD nbdcp(1) NAME nbdcp - copy between NBD servers and local files SYNOPSIS nbdcp [-a|--target-allocation allocated|sparse] [-m|--multi-conn <n>] [-M|--multi-conn-target <n>] [-p|--progress-bar] [-S|--sparse-detect <n>] [-T|--threads <n>] [-z|--target-is-zero] 'nbd://...'|DISK.IMG 'nbd://...'|DISK.IMG DESCRIPTION nbdcp is a utility that can copy quickly between NBD servers and local raw format files (or block devices). It can copy: from NBD server to file (or block device) For example, this command copies from the NBD server listening on port 10809 on "example.com" to a local file called disk.img: nbdcp nbd://example.com disk.img from file (or block device) to NBD server For example, this command copies from a local block device /dev/sda to the NBD server listening on Unix domain socket /tmp/socket: nbdcp /dev/sda 'nbd+unix:///?socket=/tmp/socket' from NBD server to NBD server For example this copies between two different exports on the same NBD server: nbdcp nbd://example.com/export1 nbd://example.com/export2 This program cannot: copy from file to file (use cp(1) or dd(1)), copy to or from formats other than raw (use qemu-img(1) convert), or access servers other than NBD servers (also use qemu-img(1)). NBD servers are specified by their URI, following the NBD URI standard at https://github.com/NetworkBlockDevice/nbd/blob/master/doc/uri.md Controlling sparseness or preallocation in the target The options -a (--target-allocation), -S (--sparse-detect) and -z (--target-is-zero) together control sparseness in the target file. By default nbdcp tries to both preserve sparseness from the source and will detect runs of allocated zeroes and turn them into sparseness. To turn off detection of sparseness use "-S 0". The -z option should be used if and only if you know that the target block device is zeroed already. This allows an important optimization where nbdcp can skip zeroing or trimming parts of the disk that are already zero. The -a option is used to control the desired final preallocation state of the target. The default is "-a sparse" which makes the target as sparse as possible. "-a allocated" makes the target fully allocated. OPTIONS --help Display brief command line help and exit. -a allocated --target-allocation=allocated Make the target fully allocated. -a sparse --target-allocation=sparse Make the target as sparse as possible. This is the default. See also "Controlling sparseness or preallocation in the target". -m N --multi-conn=N Enable NBD multi-conn with up to "N" connections. Only some NBD servers support this but it can greatly improve performance. The default is to enable multi-conn if we detect that the server supports it, with up to 4 connections. -M N --multi-conn-target=N If you are copying between NBD servers, use -m to control the multi-conn setting for the source server, and this option (-M) to control the multi-conn setting for the target server. -p --progress-bar Display a progress bar during copying. -p machine:FD --progress-bar=machine:FD Write a machine-readable progress bar to file descriptor "FD". This progress bar prints lines with the format "COPIED/TOTAL" (where "COPIED" and "TOTAL" are 64 bit unsigned integers). -S 0 --sparse-detect=0 Turn off sparseness detection. -S N --sparse-detect=N Detect runs of zero bytes of at least size "N" bytes and turn them into sparse blocks on the target (if "-a sparse" is used). This is the default, with a 512 byte block size. -T N --threads N Use at most "N" threads when copying. Usually more threads leads to better performance, up to the limit of the number of cores on your machine and the parallelism of the underlying disk or network. The default is to use the number of online processors. -z --target-is-zero Declare that the target block device contains only zero bytes (or sparseness that reads back as zeroes). You must only use this option if you are sure that this is true, since it means that nbdcp will enable an optimization where it skips zeroing parts of the disk that are zero on the source. -V --version Display the package name and version and exit. SEE ALSO qemu-img(1), libnbd(3), nbdsh(1). AUTHORS Richard W.M. Jones COPYRIGHT Copyright (C) 2020 Red Hat Inc. LICENSE This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA libnbd-1.3.1 2020-01-22 nbdcp(1)