From: Thomas Lamprecht <t.lamprecht@proxmox.com>
To: "Fabian Grünbichler" <f.gruenbichler@proxmox.com>,
"Christian Ebner" <c.ebner@proxmox.com>,
"Proxmox Backup Server development discussion"
<pbs-devel@lists.proxmox.com>
Subject: Re: [pbs-devel] [PATCH proxmox-backup] GC: s3: fix local marker cleanup for unreferenced, s3 only chunks
Date: Mon, 24 Nov 2025 10:00:25 +0100 [thread overview]
Message-ID: <bccbc898-80e2-42d3-a76f-ec3db7c1cf19@proxmox.com> (raw)
In-Reply-To: <1763972888.fq1bwvzgnc.astroid@yuna.none>
Am 24.11.25 um 09:29 schrieb Fabian Grünbichler:
> On November 24, 2025 9:22 am, Christian Ebner wrote:
>> On 11/24/25 9:13 AM, Fabian Grünbichler wrote:
>>> we basically already have such a boolean marker - we set `atime` to 0 in
>>> this case (and only this case), and we could just ignore the removal
>>> errors then? possibly limited to just ignoring ENOTFOUND?
>>
>> That's a good idea! So I will perform the additional checks based on that.
>
> alternatively, skipping cond_sweep_chunk entirely would also work (this
> is with `-w`):
>
> diff --git a/pbs-datastore/src/datastore.rs b/pbs-datastore/src/datastore.rs
> index 65299cca9..b9debd2b1 100644
> --- a/pbs-datastore/src/datastore.rs
> +++ b/pbs-datastore/src/datastore.rs
> @@ -1709,21 +1709,22 @@ impl DataStore {
> // Check local markers (created or atime updated during phase1) and
> // keep or delete chunk based on that.
> let atime = match std::fs::metadata(&chunk_path) {
> - Ok(stat) => stat.accessed()?,
> + Ok(stat) => Some(stat.accessed()?),
> Err(err) if err.kind() == std::io::ErrorKind::NotFound => {
> if self.inner.chunk_store.clear_chunk_expected_mark(&digest)? {
> unsafe {
> // chunk store lock held
> self.inner.chunk_store.replace_chunk_with_marker(&digest)?;
> }
> - SystemTime::now()
> + Some(SystemTime::now())
> } else {
> - // File not found, delete by setting atime to unix epoch
> - SystemTime::UNIX_EPOCH
> + // File not found, only delete from S3
> + None
> }
> }
> Err(err) => return Err(err.into()),
> };
> + if let Some(atime) = atime {
> let atime = atime.duration_since(SystemTime::UNIX_EPOCH)?.as_secs() as i64;
>
> unsafe {
> @@ -1752,6 +1753,13 @@ impl DataStore {
> },
> )?;
> }
> + } else {
> + // set age based on first insertion
> + if delete_list.is_empty() {
> + delete_list_age = epoch_i64();
> + }
> + delete_list.push((content.key, _chunk_guard));
> + }
This is easier to understand for me change-wise and looks OK to me (no in-depth review
though).
_______________________________________________
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:[~2025-11-24 9:00 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-11-22 10:41 Christian Ebner
2025-11-22 14:56 ` Thomas Lamprecht
2025-11-24 7:35 ` Christian Ebner
2025-11-24 8:13 ` Fabian Grünbichler
2025-11-24 8:22 ` Christian Ebner
2025-11-24 8:29 ` Fabian Grünbichler
2025-11-24 8:34 ` Christian Ebner
2025-11-24 9:00 ` Thomas Lamprecht [this message]
2025-11-24 9:41 ` Christian Ebner
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=bccbc898-80e2-42d3-a76f-ec3db7c1cf19@proxmox.com \
--to=t.lamprecht@proxmox.com \
--cc=c.ebner@proxmox.com \
--cc=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