From 578f7972f273f6ca5756e66f0c9f103e44ae4eac Mon Sep 17 00:00:00 2001
From: David Alves <david.alves@dfinity.org>
Date: Tue, 22 Aug 2023 15:01:10 -0700
Subject: [PATCH] Allow to obtain the number of pending checkpoints

This is a measure of backpressure: if the number of checkpoints
pendings increases too much it means we're falling behind.

This also make it so that we panic!() when a checkpoint fails.
It's important to fail here because most of the time we don't
know how to recover from a checkpoint having failed.
---
 src/lib.rs | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/src/lib.rs b/src/lib.rs
index a5bc80a..af309fa 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -269,6 +269,12 @@ where
         }
     }
 
+    /// Returns the number of checkpoints that are pending.
+    #[must_use]
+    pub fn pending_checkpoints(&self) -> usize {
+        self.data.checkpoint_sender.len()
+    }
+
     fn reclaim(
         &self,
         file: LogFile<M::File>,
@@ -359,8 +365,11 @@ where
                 drop(writer);
                 let mut manager = wal.data.manager.lock();
                 if let Err(error) = manager.checkpoint_to(entry_id, &mut reader, &wal) {
-                    error!("Checkpointer failed with error: {error:?}. Skipping checkpoint");
-                    continue;
+                    let message = format!(
+                        "Fatal Error: Checkpointer failed with error: {error:?}. Cannot proceed."
+                    );
+                    error!("{}", message);
+                    panic!("{}", message);
                 }
                 writer = file_to_checkpoint.lock();
                 Some(entry_id)