Re: Fix lag columns in pg_stat_replication not advancing when replay LSN stalls

> On Oct 17, 2025, at 11:56, Fujii Masao <masao.fujii@gmail.com> wrote:
>
> Hi,
>
> While testing, I noticed that write_lag and flush_lag in pg_stat_replication
> initially advanced but eventually stopped updating. This happened when
> I started pg_receivewal, ran pgbench, and periodically monitored
> pg_stat_replication.
>
> My analysis shows that this issue occurs when any of the write, flush,
> or replay LSNs in the standby’s feedback message stop updating for some time.
> In the case of pg_receivewal, the replay LSN is always invalid (never updated),
> which triggers the problem. Similarly, in regular streaming replication,
> if the replay LSN remains unchanged for a long time—such as during
> a recovery conflict—the lag values for both write and flush can stop advancing.
>
> The root cause seems to be that when any of the LSNs stop updating,
> the lag tracker's cyclic buffer becomes full (the write head reaches
> the slowest read head). In this situation, LagTrackerWrite() and
> LagTrackerRead() didn't handle the full-buffer condition properly.
> For instance, if the replay LSN stalls, the buffer fills up and the read heads
> for "write" and "flush" end up at the same position as the write head.
> This causes LagTrackerRead() to return -1 for both, preventing write_lag
> and flush_lag from advancing.
>
> The attached patch fixes the problem by treating the slowest read entry
> (the one causing the buffer to fill up) as a separate overflow entry,
> allowing the lag tracker to continue operating correctly.
>
> --
> Fujii Masao
> <v1-0001-Fix-lag-columns-in-pg_stat_replication-not-advanc.patch>

It took me some time to understand this fix. My most confusing was that once overwrite happens, how a reader head to
catchup again? Finally I figured it out: 

```
+        lag_tracker->read_heads[head] =
+            (lag_tracker->write_head + 1) % LAG_TRACKER_BUFFER_SIZE;
```

"(lag_tracker->write_head + 1) % LAG_TRACKER_BUFFER_SIZE” points to the oldest LSN in the ring, from where an
overflowedreader head starts to catch up. 

I have no comment on the code change. Nice patch!

All I wonder is if we can add a TAP test for this fix?

Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/