On Tue, Feb 18, 2025 at 03:52:26PM +0100, Gabriel Goller wrote:
> The workertasks currently get their status from parsing the log
> messages in the task-log file. The problem is that if these messages are
> filtered – which is now possible using the PBS_LOG env variable – some
> workertasks will end up with a "stopped: unknown" status. This is not
> desirable so write the message manually to the workertask file and
> bypass tracing.
> 
> This way we are guaranteed that, regardless of the max logging level the
> user sets, the final message (and status) is written.
> 
> Signed-off-by: Gabriel Goller <g.gol...@proxmox.com>
> ---
> 
> Changelog:
> v2:
>  * remove jank (detect workertask failure based on return value instead
>    of message)
> 
>  proxmox-log/src/lib.rs                 | 17 +++++++++++++++++
>  proxmox-rest-server/src/worker_task.rs | 16 +++++++++++++---
>  2 files changed, 30 insertions(+), 3 deletions(-)
> 
> diff --git a/proxmox-log/src/lib.rs b/proxmox-log/src/lib.rs
> index 8c74e42b618d..755d1b4a850c 100644
> --- a/proxmox-log/src/lib.rs
> +++ b/proxmox-log/src/lib.rs
> @@ -191,3 +191,20 @@ pub fn init_cli_logger(
>      LogTracer::init_with_filter(log_level.as_log())?;
>      Ok(())
>  }
> +
> +/// Write manually to the current tasklog bypassing the whole tracing 
> infrastructure. Note that this
> +/// will also bypass all the filtering and writing to journald or elsewhere. 
> If has_failed is true,
> +/// print to stderr as well.
> +pub fn log_manually_to_tasklog(msg: &str, has_failed: bool) -> Result<(), 
> anyhow::Error> {

What's "manual" about using a provided helper function? :-P
Besides, the name kind of conflicts with the stderr write, which seems
even more specific to the rest-server case.

Maybe a `LogContext::log_unfiltered(&str)` and rest-server just calls
this+eprintln!()? Then the `error!()` invocation could be in a `None`
match arm on the `LogContext::current()` match which IMO makes for much
nicer control flow.

Do we even anticipate any other use case than the one in rest-server?

> +    if let Some(ctx) = LogContext::current() {
> +        let mut logger = ctx.logger.lock().unwrap();
> +        // If the message is an error, print to stderr (which will end up in 
> journald) as well.
> +        if has_failed {
> +            eprintln!("{msg}");
> +        }
> +        logger.logger.log(msg);
> +        Ok(())
> +    } else {
> +        anyhow::bail!("Unable to manually write message to workertask log 
> file: No workertask exists in this task.");
> +    }
> +}
> diff --git a/proxmox-rest-server/src/worker_task.rs 
> b/proxmox-rest-server/src/worker_task.rs
> index 24e2676e69c7..06d986f41dbb 100644
> --- a/proxmox-rest-server/src/worker_task.rs
> +++ b/proxmox-rest-server/src/worker_task.rs
> @@ -7,14 +7,14 @@ use std::sync::atomic::{AtomicBool, AtomicUsize, Ordering};
>  use std::sync::{Arc, LazyLock, Mutex, OnceLock};
>  use std::time::{Duration, SystemTime};
>  
> -use anyhow::{bail, format_err, Error};
> +use anyhow::{bail, format_err, Context, Error};
>  use futures::*;
>  use nix::fcntl::OFlag;
>  use serde::{Deserialize, Serialize};
>  use serde_json::{json, Value};
>  use tokio::signal::unix::SignalKind;
>  use tokio::sync::{oneshot, watch};
> -use tracing::{info, warn};
> +use tracing::{error, info, warn};
>  
>  use proxmox_daemon::command_socket::CommandSocket;
>  use proxmox_lang::try_block;
> @@ -1025,7 +1025,17 @@ impl WorkerTask {
>      /// Log task result, remove task from running list
>      pub fn log_result(&self, result: &Result<(), Error>) {
>          let state = self.create_state(result);
> -        self.log_message(state.result_text());
> +
> +        // Write the result manually to the workertask file. We don't want 
> to filter or process
> +        // this message by the logging system. This also guarantees the 
> result message will be in
> +        // the file, regardless of the logging level.
> +        proxmox_log::log_manually_to_tasklog(&state.result_text(), 
> result.is_err())
> +            .context("Error writing final result message")
> +            .unwrap_or_else(|err| {
> +                // Note: this will probably also fail to be written to the 
> tasklog, but at least it
> +                // ends up in the journal.
> +                error!("{err:#}");
> +            });
>  
>          WORKER_TASK_LIST.lock().unwrap().remove(&self.upid.task_id);
>          // this wants to access WORKER_TASK_LIST, so we need to drop the 
> lock above
> -- 
> 2.39.5
> 
> 
> 
> _______________________________________________
> pve-devel mailing list
> pve-devel@lists.proxmox.com
> https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel


_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel

Reply via email to