FATAL","57M02","the database system is in recovery mode","last replayed record at 0/C39A3F8 on all mirror segment logs.

Products

• VMware Tanzu Greenplum

Issue/Introduction

FATAL error in all the mirrors logs on his 6.x system
 

FATAL","57M02","the database system is in recovery mode","last replayed record at 0/C39A3F8

Also if you run  a check of the processes on the segment hosts, each mirror, has a  startup recovery process

 

[gpadmin@p1sdw1 ~]$ ps -ef |grep -i recover 

gpadmin 283166 283044 0 2020 ? 08:14:54 postgres: 5001, startup process recovering 000000010000088100000016 
gpadmin 283169 283042 0 2020 ? 08:33:37 postgres: 5000, startup process recovering 00000001000008820000002D 
gpadmin 283174 283041 0 2020 ? 08:32:03 postgres: 5002, startup process recovering 000000010000088C00000025 
gpadmin 283191 283028 0 2020 ? 08:22:34 postgres: 5023, startup process recovering 000000010000085700000019 
gpadmin 283192 283027 0 2020 ? 08:11:04 postgres: 5019, startup process recovering 000000030000085300000032 
gpadmin 283196 283001 0 2020 ? 07:53:21 postgres: 5006, startup process recovering 00000001000008810000002E 
gpadmin 283200 283026 0 2020 ? 08:38:41 postgres: 5021, startup process recovering 000000010000085900000009

Environment

Product Version: 6.5

Resolution

This is functioning as designed and not a cause of concern. These messages are present as we changed to WAL replication for primary to mirror replication. Normally, we can safely ignore this error. However, if you have further concerns, please get in touch with Broadcom Support for assistance.