mirror of
https://github.com/openai/codex.git
synced 2026-11-13 21:30:47 +00:00
Anthony Tafoya
3fd08b5405
## Why Failed cloud discovery previously retried before each model step. Apply a cooldown across steps and turns to avoid repeated discovery requests while a provider remains unavailable. ## What changed - Share per-thread retry backoff logic between cloud skill and plugin catalogs, starting at 30 seconds and doubling to a maximum of 5 minutes. - Scope cooldowns to authentication and MCP resource generations so account changes and reconnects permit immediate recovery. - Clear backoff after successful discovery or configuration resets, while allowing cancelled refreshes to retry immediately. ## Testing Update lifecycle and concurrency tests to cover shared cooldowns, reconnect recovery, and cancellation. Use paused time to verify exponential delays, the 5-minute cap, and backoff across turns. GitOrigin-RevId: 30dab5ceba51e8ad64c0794c1729719e99f577ce
56 lines
1.7 KiB
Rust
56 lines
1.7 KiB
Rust
//! Thread-owned cloud catalog retry cooldowns scoped to auth and resource generations.
|
|
|
|
use codex_mcp::McpResourceClientAuthKey;
|
|
use codex_mcp::McpResourceServerCacheKey;
|
|
use std::time::Duration;
|
|
use tokio::time::Instant;
|
|
|
|
/// Store per catalog and thread; clear on success or configuration reset.
|
|
#[derive(Clone)]
|
|
pub struct CatalogRefreshBackoff {
|
|
auth: Option<McpResourceClientAuthKey>,
|
|
resources: Option<McpResourceServerCacheKey>,
|
|
delay: Duration,
|
|
retry_at: Instant,
|
|
}
|
|
|
|
impl std::fmt::Debug for CatalogRefreshBackoff {
|
|
fn fmt(&self, formatter: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
|
|
formatter
|
|
.debug_struct("CatalogRefreshBackoff")
|
|
.field("delay", &self.delay)
|
|
.field("retry_at", &self.retry_at)
|
|
.finish_non_exhaustive()
|
|
}
|
|
}
|
|
|
|
impl CatalogRefreshBackoff {
|
|
pub fn is_pending(
|
|
&self,
|
|
auth: Option<&McpResourceClientAuthKey>,
|
|
resources: Option<&McpResourceServerCacheKey>,
|
|
) -> bool {
|
|
self.auth.as_ref() == auth
|
|
&& self.resources.as_ref() == resources
|
|
&& Instant::now() < self.retry_at
|
|
}
|
|
|
|
pub fn after_failure(
|
|
auth: Option<McpResourceClientAuthKey>,
|
|
resources: Option<McpResourceServerCacheKey>,
|
|
previous: Option<&Self>,
|
|
) -> Self {
|
|
let delay = previous
|
|
.filter(|previous| previous.auth == auth && previous.resources == resources)
|
|
.map_or(Duration::from_secs(30), |previous| {
|
|
(previous.delay * 2).min(Duration::from_secs(300))
|
|
});
|
|
Self {
|
|
auth,
|
|
resources,
|
|
delay,
|
|
retry_at: Instant::now() + delay,
|
|
}
|
|
}
|
|
}
|