From: "Fabian Grünbichler" <f.gruenbichler@proxmox.com>
To: Proxmox Backup Server development discussion
<pbs-devel@lists.proxmox.com>
Subject: Re: [pbs-devel] [PATCH vma-to-pbs v4 2/6] add support for bulk import of a dump directory
Date: Mon, 04 Nov 2024 14:06:13 +0100 [thread overview]
Message-ID: <1730724842.agk2is6zq8.astroid@yuna.none> (raw)
In-Reply-To: <20241030135537.92595-3-f.schauer@proxmox.com>
On October 30, 2024 2:55 pm, Filip Schauer wrote:
> When a path to a directory is provided in the vma_file argument, try to
> upload all VMA backups in the directory. This also handles compressed
> VMA files, notes and logs. If a vmid is specified with --vmid, only the
> backups of that particular vmid are uploaded.
>
> This is intended for use on a dump directory:
>
> PBS_FINGERPRINT='PBS_FINGERPRINT' vma-to-pbs \
> --repository 'user@realm!token@server:port:datastore' \
> /var/lib/vz/dump
>
> Signed-off-by: Filip Schauer <f.schauer@proxmox.com>
> ---
> Cargo.toml | 3 +
> src/main.rs | 161 +++++++++++++++++++++++++++++++++++++++++++++----
> src/vma2pbs.rs | 64 +++++++++++++++++---
> 3 files changed, 209 insertions(+), 19 deletions(-)
>
> diff --git a/Cargo.toml b/Cargo.toml
> index cd13426..5c6a175 100644
> --- a/Cargo.toml
> +++ b/Cargo.toml
> @@ -7,9 +7,12 @@ edition = "2021"
> [dependencies]
> anyhow = "1.0"
> bincode = "1.3"
> +chrono = "0.4"
> hyper = "0.14.5"
> +itertools = "0.13"
not needed, see below
> pico-args = "0.5"
> md5 = "0.7.0"
> +regex = "1.7"
> scopeguard = "1.1.0"
> serde = "1.0"
> serde_json = "1.0"
> diff --git a/src/main.rs b/src/main.rs
> index 3e25591..4c5135b 100644
> --- a/src/main.rs
> +++ b/src/main.rs
> @@ -1,26 +1,35 @@
> use std::ffi::OsString;
> +use std::fs::read_dir;
> +use std::io::{BufRead, BufReader};
> +use std::path::PathBuf;
>
> use anyhow::{bail, Context, Error};
> +use chrono::NaiveDateTime;
> +use itertools::Itertools;
> use proxmox_sys::linux::tty;
> use proxmox_time::epoch_i64;
> +use regex::Regex;
>
> mod vma;
> mod vma2pbs;
> -use vma2pbs::{vma2pbs, BackupVmaToPbsArgs, PbsArgs, VmaBackupArgs};
> +use vma2pbs::{vma2pbs, BackupVmaToPbsArgs, Compression, PbsArgs, VmaBackupArgs};
>
> const CMD_HELP: &str = "\
> Usage: vma-to-pbs [OPTIONS] --repository <auth_id@host:port:datastore> --vmid <VMID> [vma_file]
>
> Arguments:
> - [vma_file]
> + [vma_file | dump_directory]
>
> Options:
> --repository <auth_id@host:port:datastore>
> Repository URL
> [--ns <NAMESPACE>]
> Namespace
> - --vmid <VMID>
> + [--vmid <VMID>]
> Backup ID
> + This is required if a single VMA file is provided.
> + If not specified, bulk import all VMA backups in the provided directory.
> + If specified with a dump directory, only import backups of the specified vmid.
> [--backup-time <EPOCH>]
> Backup timestamp
> --fingerprint <FINGERPRINT>
> @@ -41,6 +50,8 @@ Options:
> File containing a comment/notes
> [--log-file <LOG_FILE>]
> Log file
> + -y, --yes
> + Automatic yes to prompts
> -h, --help
> Print help
> -V, --version
> @@ -52,7 +63,16 @@ fn parse_args() -> Result<BackupVmaToPbsArgs, Error> {
> args.remove(0); // remove the executable path.
>
> let mut first_later_args_index = 0;
> - let options = ["-h", "--help", "-c", "--compress", "-e", "--encrypt"];
> + let options = [
> + "-h",
> + "--help",
> + "-c",
> + "--compress",
> + "-e",
> + "--encrypt",
> + "-y",
> + "--yes",
> + ];
>
> for (i, arg) in args.iter().enumerate() {
> if let Some(arg) = arg.to_str() {
> @@ -87,7 +107,7 @@ fn parse_args() -> Result<BackupVmaToPbsArgs, Error> {
>
> let pbs_repository = args.value_from_str("--repository")?;
> let namespace = args.opt_value_from_str("--ns")?;
> - let vmid = args.value_from_str("--vmid")?;
> + let vmid = args.opt_value_from_str("--vmid")?;
> let backup_time: Option<i64> = args.opt_value_from_str("--backup-time")?;
> let backup_time = backup_time.unwrap_or_else(epoch_i64);
> let fingerprint = args.opt_value_from_str("--fingerprint")?;
> @@ -99,6 +119,7 @@ fn parse_args() -> Result<BackupVmaToPbsArgs, Error> {
> let key_password_file: Option<OsString> = args.opt_value_from_str("--key-password-file")?;
> let notes_file: Option<OsString> = args.opt_value_from_str("--notes-file")?;
> let log_file_path: Option<OsString> = args.opt_value_from_str("--log-file")?;
> + let yes = args.contains(["-y", "--yes"]);
>
> match (encrypt, keyfile.is_some()) {
> (true, false) => bail!("--encrypt requires a --keyfile!"),
> @@ -196,15 +217,131 @@ fn parse_args() -> Result<BackupVmaToPbsArgs, Error> {
> encrypt,
> };
>
> - let vma_args = VmaBackupArgs {
> - vma_file_path: vma_file_path.cloned(),
> - backup_id: vmid,
> - backup_time,
> - notes,
> - log_file_path,
> + let bulk =
> + vma_file_path
> + .map(PathBuf::from)
> + .and_then(|path| if path.is_dir() { Some(path) } else { None });
> +
> + let grouped_vmas = if let Some(dump_dir_path) = bulk {
grouped_vmas should still be a map, not a vec of vec..
e.g., something like this (requires some more adaptation - while this
could use itertools, I don't think it's worth to pull that in if the
same can be had with a single fold invocation):
@@ -298,12 +298,16 @@ fn parse_args() -> Result<BackupVmaToPbsArgs, Error> {
vmas.sort_by_key(|d| d.backup_time);
let total_vma_count = vmas.len();
- let mut grouped_vmas: Vec<_> = vmas
- .into_iter()
- .into_group_map_by(|d| d.backup_id.clone())
- .into_values()
- .collect();
- grouped_vmas.sort_by_key(|d| d[0].backup_id.clone());
+ let grouped_vmas = vmas.into_iter().fold(
+ HashMap::new(),
+ |mut grouped: HashMap<String, Vec<VmaBackupArgs>>, vma_args| {
+ grouped
+ .entry(vma_args.backup_id.clone())
+ .or_default()
+ .push(vma_args);
+ grouped
+ },
+ );
log::info!(
"Found {} backup archive(s) of {} different VMID(s):",
@@ -311,12 +315,8 @@ fn parse_args() -> Result<BackupVmaToPbsArgs, Error> {
grouped_vmas.len()
);
- for vma_group in &grouped_vmas {
- log::info!(
- "- VMID {}: {} backups",
- vma_group[0].backup_id,
- vma_group.len()
- );
+ for (vma_group, vma_args) in &grouped_vmas {
+ log::info!("- VMID {}: {} backups", vma_group, vma_args.len());
}
if !yes {
> + let re = Regex::new(
> + r"vzdump-qemu-(\d+)-(\d{4}_\d{2}_\d{2}-\d{2}_\d{2}_\d{2}).vma(|.zst|.lzo|.gz)$",
> + )?;
> +
> + let mut vmas = Vec::new();
> +
> + for entry in read_dir(dump_dir_path)? {
> + let entry = entry?;
> + let path = entry.path();
> +
> + if !path.is_file() {
> + continue;
> + }
> +
> + if let Some(file_name) = path.file_name().and_then(|n| n.to_str()) {
> + let Some((_, [backup_id, timestr, ext])) =
> + re.captures(file_name).map(|c| c.extract())
> + else {
> + // Skip the file, since it is not a VMA backup
> + continue;
> + };
> +
> + if let Some(ref vmid) = vmid {
> + if backup_id != vmid {
> + // Skip the backup, since it does not match the specified vmid
> + continue;
> + }
> + }
> +
> + let compression = match ext {
> + "" => None,
> + ".zst" => Some(Compression::Zstd),
> + ".lzo" => Some(Compression::Lzo),
> + ".gz" => Some(Compression::GZip),
> + _ => bail!("Unexpected file extension: {ext}"),
> + };
> +
> + let backup_time = NaiveDateTime::parse_from_str(timestr, "%Y_%m_%d-%H_%M_%S")?
> + .and_utc()
> + .timestamp();
> +
> + let notes_path = path.with_file_name(format!("{}.notes", file_name));
> + let notes = proxmox_sys::fs::file_read_optional_string(notes_path)?;
> +
> + let log_path = path.with_file_name(format!("{}.log", file_name));
> + let log_file_path = if log_path.exists() {
> + Some(log_path.to_path_buf().into_os_string())
> + } else {
> + None
> + };
> +
> + let backup_args = VmaBackupArgs {
> + vma_file_path: Some(path.clone().into()),
> + compression,
> + backup_id: backup_id.to_string(),
> + backup_time,
> + notes,
> + log_file_path,
> + };
> + vmas.push(backup_args);
> + }
> + }
> +
> + vmas.sort_by_key(|d| d.backup_time);
> + let total_vma_count = vmas.len();
> + let mut grouped_vmas: Vec<_> = vmas
> + .into_iter()
> + .into_group_map_by(|d| d.backup_id.clone())
> + .into_values()
> + .collect();
> + grouped_vmas.sort_by_key(|d| d[0].backup_id.clone());
> +
> + println!(
> + "Found {} backup archive(s) of {} different VMID(s):",
> + total_vma_count,
> + grouped_vmas.len()
> + );
if we don't find any, we should print something else here and exit?
> +
> + for vma_group in &grouped_vmas {
> + println!(
> + "- VMID {}: {} backups",
> + vma_group[0].backup_id,
> + vma_group.len()
> + );
> + }
> +
> + if !yes {
> + loop {
> + eprint!("Proceed with the bulk import? (y/n): ");
> + let mut line = String::new();
> +
> + BufReader::new(std::io::stdin()).read_line(&mut line)?;
> + let trimmed = line.trim();
> + if trimmed == "y" || trimmed == "Y" {
> + break;
> + } else if trimmed == "n" || trimmed == "N" {
> + bail!("Bulk import was not confirmed.");
> + }
this maybe should mimic what we do in proxmox_router when prompting for
confirmation? e.g., flush stdout, have a default value, ..?
should we abort after a few loops?
> + }
> + }
> +
> + grouped_vmas
_______________________________________________
pbs-devel mailing list
pbs-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pbs-devel
next prev parent reply other threads:[~2024-11-04 13:06 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-30 13:55 [pbs-devel] [PATCH vma-to-pbs v4 0/6] " Filip Schauer
2024-10-30 13:55 ` [pbs-devel] [PATCH vma-to-pbs v4 1/6] split BackupVmaToPbsArgs into PbsArgs and VmaBackupArgs Filip Schauer
2024-10-30 13:55 ` [pbs-devel] [PATCH vma-to-pbs v4 2/6] add support for bulk import of a dump directory Filip Schauer
2024-11-04 13:06 ` Fabian Grünbichler [this message]
2024-10-30 13:55 ` [pbs-devel] [PATCH vma-to-pbs v4 3/6] add option to skip vmids whose backups failed to upload Filip Schauer
2024-10-30 13:55 ` [pbs-devel] [PATCH vma-to-pbs v4 4/6] remove hard coded values Filip Schauer
2024-10-30 13:55 ` [pbs-devel] [PATCH vma-to-pbs v4 5/6] use level-based logging instead of println Filip Schauer
2024-10-30 13:55 ` [pbs-devel] [PATCH vma-to-pbs v4 6/6] log device upload progress as a percentage Filip Schauer
2024-11-04 13:09 ` [pbs-devel] partially-applied: [PATCH vma-to-pbs v4 0/6] add support for bulk import of a dump directory Fabian Grünbichler
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1730724842.agk2is6zq8.astroid@yuna.none \
--to=f.gruenbichler@proxmox.com \
--cc=pbs-devel@lists.proxmox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox