From f188bdef53ee8c71ec4070172d7b186a03fcd04a Mon Sep 17 00:00:00 2001
From: Jakov Sosic <jsosic@gmail.com>
Date: Thu, 24 Dec 2020 16:39:43 +0100
Subject: [PATCH] Improve PostgreSQL replication lag detection (#395)

In some cases master can show pg_last_xact_replay_timestamp() from past,
which can cause the exporter to show ever-growing value for the lag.

By checking if the instance is in recovery we can avoid reporting some
huge number for master instance.
---
 queries.yaml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/queries.yaml b/queries.yaml
index 24abb9a9..800e62bd 100644
--- a/queries.yaml
+++ b/queries.yaml
@@ -1,5 +1,5 @@
 pg_replication:
-  query: "SELECT EXTRACT(EPOCH FROM (now() - pg_last_xact_replay_timestamp())) as lag"
+  query: "SELECT CASE WHEN NOT pg_is_in_recovery() THEN 0 ELSE GREATEST (0, EXTRACT(EPOCH FROM (now() - pg_last_xact_replay_timestamp()))) END AS lag"
   master: true
   metrics:
     - lag: