From: Manuel Federanko <m.federanko@proxmox.com>
To: Shan Shaji <s.shaji@proxmox.com>,
pbs-devel@lists.proxmox.com, pdm-devel@lists.proxmox.com
Subject: Re: [PATCH proxmox-backup 6/7] acme: fix #6372 implement ARI renewal information fetching.
Date: Wed, 29 Jul 2026 11:08:26 +0200 [thread overview]
Message-ID: <3f6adb7e-3432-436d-a0d2-d2bac8959196@proxmox.com> (raw)
In-Reply-To: <DKA8QK61HAXK.33U6CVIMUD23T@proxmox.com>
On 2026-07-28 3:32 PM, Shan Shaji wrote:
> [snip]
>
>> +/// ARI renewal time if available
>> +///
>> +/// Query the ARI endpoint for a suggested renewal window, draw a uniform random time in this window
>> +/// Return None if ARI does not apply.
>> +async fn cert_renew_lead_time_ari(
>> + acme_config: &AcmeConfig,
>> + cert_info: &cert::CertInfo,
>> +) -> Result<Option<i64>, Error> {
>> + let now = proxmox_time::epoch_i64();
>> + if cert_info.is_expired_after_epoch(now)? {
>> + return Ok(Some(0));
>> + }
>> + let ari_id = &match cert_info.ari_id() {
>
> tiny nit: I think this `&` is not necessary here as deref coercion happens later.
ack
>> + Some(x) => x,
>> + None => return Ok(None),
>> + };
>> + let window = match proxmox_acme_api::get_renewal_info(acme_config, &ari_id).await? {
>> + Some(x) => x,
>> + None => return Ok(None),
>> + };
>> + if let Some(reason) = window.data.explanation_url {
>> + info!(
>> + "Obtained renewal window, for information on this chosen window please visit {reason}"
>> + );
>> + }
>> + let window_start = proxmox_time::parse_rfc3339(&window.data.suggested_window.start)?;
>> + let window_end = proxmox_time::parse_rfc3339(&window.data.suggested_window.end)?;
>> + let rand = proxmox_sys::linux::random_data(8)?
>> + .into_iter()
>> + .enumerate()
>> + .fold(0, |acc, (index, x)| acc + ((x as u64) << (index * 8))) as f64
>
> optional: Instead of manually creating u64 from the bytes, would it be possible to use
> `u64::from_le_bytes` here?
>
> May be like: **untested**
>
> let mut buffer = [0u8; 8];
> proxmox_sys::linux::fill_with_random_data(&mut buffer)?;
> let rand = u64::from_le_bytes(buffer) as f64 / u64::MAX as f64;
yeah that makes sense
>> + / (u64::MAX as f64);
>> + let renew = window_start + (((window_end - window_start) as f64) * rand) as i64;
>> + // need max since the randomness could result in negative values
>> + Ok(Some(std::cmp::max(0, renew - now)))
>> +}
>> +
>> /// Check whether the current certificate expires within its renewal lead time.
>> ///
>> /// Returns `(expires_soon, lead_time_in_days)`; the lead time is returned so callers can produce
>> /// consistent user-facing messages without re-reading and re-parsing the certificate.
>
> nit: Since the return type is now updated, I think we should update this doc comment as well.
ack, thanks for pointing that out.
>> -pub fn check_renewal_needed() -> Result<(bool, i64), Error> {
>> - let cert = pem_to_cert_info(get_certificate_pem()?.as_bytes())?;
>> - let lead = cert_renew_lead_time(&cert);
>> - let expires_soon = cert
>> - .is_expired_after_epoch(proxmox_time::epoch_i64() + lead)
>> - .map_err(|err| format_err!("Failed to check certificate expiration date: {}", err))?;
>> - Ok((expires_soon, lead / SECONDS_PER_DAY))
>> +async fn check_renewal_needed(
>> + acme_config: &AcmeConfig,
>> + cert_info: &cert::CertInfo,
>> +) -> Result<bool, Error> {
>> + let lead_ari = cert_renew_lead_time_ari(acme_config, cert_info).await?;
>> + if let Some(lead_ari) = lead_ari {
>> + // rfc9773 section 4.2 tells us to renew if the chosen renewal time would be before the next check
>> + let expires_soon = lead_ari < SECONDS_PER_DAY;
>> + let ts = proxmox_time::TimeSpan::from(std::time::Duration::new(lead_ari as u64, 0));
>> + info!("Certificate is scheduled for renewal in {ts:.0} by ARI");
>> + Ok(expires_soon)
>> + } else {
>> + let lead = cert_renew_lead_time(&cert_info);
>> + let expires_soon = cert_info
>> + .is_expired_after_epoch(proxmox_time::epoch_i64() + lead)
>> + .map_err(|err| format_err!("Failed to check certificate expiration date: {}", err))?;
>> + let ts = proxmox_time::TimeSpan::from(std::time::Duration::new(lead as u64, 0));
>> + info!("Certificate renewal lead time is {ts:.0}");
>> + Ok(expires_soon)
>> + }
>> }
>>
>
> [snip]
>
>> --
>> 2.47.3
>
next prev parent reply other threads:[~2026-07-29 9:09 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-25 14:13 [PATCH proxmox{,-backup,-datacenter-manager} 0/7] acme: fix #6372 implement basic ARI support Manuel Federanko
2026-06-25 14:13 ` [PATCH proxmox 1/7] acme: client: add methods to fetch renewal information Manuel Federanko
2026-07-24 11:16 ` Shan Shaji
2026-07-29 9:04 ` Manuel Federanko
2026-06-25 14:13 ` [PATCH proxmox 2/7] acme: add retry-after header to " Manuel Federanko
2026-06-25 14:13 ` [PATCH proxmox 3/7] acme: allow specifying the certificate that is replaced by an order Manuel Federanko
2026-06-25 14:13 ` [PATCH proxmox 4/7] acme: cert: add dedicated ari_id field to the certificate info Manuel Federanko
2026-06-25 14:13 ` [PATCH proxmox-backup 5/7] acme: add ari_id to cert info Manuel Federanko
2026-06-25 14:13 ` [PATCH proxmox-backup 6/7] acme: fix #6372 implement ARI renewal information fetching Manuel Federanko
2026-07-27 13:31 ` Shan Shaji
2026-07-29 9:07 ` Manuel Federanko
2026-07-30 15:09 ` Shan Shaji
2026-07-28 13:32 ` Shan Shaji
2026-07-29 9:08 ` Manuel Federanko [this message]
2026-06-25 14:13 ` [PATCH proxmox-datacenter-manager 7/7] acme: fix #6372 use ARI for renewal if available Manuel Federanko
2026-07-27 16:15 ` [PATCH proxmox{,-backup,-datacenter-manager} 0/7] acme: fix #6372 implement basic ARI support Shan Shaji
2026-07-29 9:01 ` Manuel Federanko
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=3f6adb7e-3432-436d-a0d2-d2bac8959196@proxmox.com \
--to=m.federanko@proxmox.com \
--cc=pbs-devel@lists.proxmox.com \
--cc=pdm-devel@lists.proxmox.com \
--cc=s.shaji@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