Re: PostgreSQL 14 standby recovery stuck in archive recovery state

" The startup process continues to replay WAL records, but progress appears
to stop."

Can you confirm the necessary WAL files are actually available? How does
your "restore_command" retrieve them, and are the files actually there?

On Fri, Aug 7, 2026 at 6:56 PM Vijay Raut <vjraaut(at)gmail(dot)com> wrote:

> Hello PostgreSQL Community,
>
> We are facing an issue with a PostgreSQL 14 standby/replica instance where
> recovery appears to be stuck and the database is not becoming available for
> read connections.
> Environment
>
> -
>
> PostgreSQL Version: 14.23
> -
>
> OS: Ubuntu 22.04
> -
>
> Architecture: x86_64
> -
>
> Deployment: Streaming replication standby
> -
>
> Data directory:
>
> /data/pgdata_14
>
>
> Issue Description
>
> The replica started recovery after an interruption. PostgreSQL logs show:
>
> LOG: database system was interrupted while in recovery at log time 2026-08-06 05:26:34 IST
> LOG: entering standby mode
> LOG: redo starts at 13B9/6D355F08
>
> Recovery starts successfully, but the instance remains in recovery and
> does not reach a consistent state where connections are accepted.
>
> Connection attempts fail with:
>
> FATAL: the database system is not yet accepting connections
> DETAIL: Consistent recovery state has not been yet reached.
>
> The startup process continues to replay WAL records, but progress appears
> to stop.
> PostgreSQL pg_controldata output
>
> pg_control version number: 1300
> Catalog version number: 202107181
> Database system identifier: 7479073413122508773
>
> Database cluster state: in archive recovery
>
> pg_control last modified: Fri 07 Aug 2026 11:50:35 PM IST
>
> Latest checkpoint location: 13B9/73735BC8
> Latest checkpoint's REDO location: 13B9/6D355F08
> Latest checkpoint's REDO WAL file: 0000000F000013B90000006D
>
> Latest checkpoint's TimeLineID: 15
> Latest checkpoint's PrevTimelineID: 15
>
> Minimum recovery ending location: 13BA/10996458
> Min recovery ending loc's timeline: 15
>
> Backup start location: 13AC/C59613E8
> Backup end location: 0/0
> End-of-backup record required: yes
>
> wal_level setting: replica
> wal_log_hints setting: on
>
> max_connections: 2000
> max_worker_processes: 16
> max_wal_senders: 10
>
> Data page checksum version: 0
>
> Additional observations
>
> The PostgreSQL log contains many WAL redo messages similar to:
>
> DEBUG: WAL redo at 13B9/...
>
> The startup process is active, but recovery does not complete.
>
> We checked storage-related issues:
>
> -
>
> /data filesystem is healthy from space perspective.
> -
>
> PostgreSQL data directory is on:
>
> /dev/mapper/VG01-data_lv
>
> -
>
> No clear filesystem I/O errors were observed.
> -
>
> Kernel logs showed previous DISCARD/TRIM related messages:
>
> blk_update_request: critical target error, dev sdb, sector xxxx op 0x3:(DISCARD)
>
> but no confirmed read/write I/O failures.
> Checks performed
>
> 1.
>
> PostgreSQL process status:
>
> ps -ef | grep postgres
>
> Questions
>
> 1.
>
> What could prevent PostgreSQL 14 standby recovery from reaching a
> consistent recovery state when WAL replay appears active?
> 2.
>
> Is there a way to identify which WAL record or relation is preventing
> recovery completion?
>
> Any guidance on further troubleshooting would be appreciated.
>
> Regards,
> Vijay R
>