From 945f3b50fc19be60fe81244916897ba736714ea9 Mon Sep 17 00:00:00 2001
From: Christian Schwarz <christian@neon.tech>
Date: Sat, 23 Mar 2024 13:48:00 +0000
Subject: [PATCH] walreceiver-after-ingest failpoint: use pausable_failpoint

Possibly the last commit was a red herring (still a good cleanup though).

However, this one here is a definite executor stall at the failpoint.

And it makes sense that it would cause the shutdown timeouts with the
unified runtime, because the walreceiver code previously ran on a
separate runtime.
---
 .../src/tenant/timeline/walreceiver/walreceiver_connection.rs  | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs b/pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs
index 784c4287de6fe..459c117f66801 100644
--- a/pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs
+++ b/pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs
@@ -11,7 +11,6 @@ use std::{
 use anyhow::{anyhow, Context};
 use bytes::BytesMut;
 use chrono::{NaiveDateTime, Utc};
-use fail::fail_point;
 use futures::StreamExt;
 use postgres::{error::SqlState, SimpleQueryMessage, SimpleQueryRow};
 use postgres_ffi::WAL_SEGMENT_SIZE;
@@ -326,7 +325,7 @@ pub(super) async fn handle_walreceiver_connection(
                             filtered_records += 1;
                         }
 
-                        fail_point!("walreceiver-after-ingest");
+                        pausable_failpoint!("walreceiver-after-ingest");
 
                         last_rec_lsn = lsn;