Add arch_dest config param when use olr in DataGard

[zwjcool]

Is your feature request related to a problem? Please describe.
Hi, I get error follows:
2025-06-23 14:57:43 ERROR 10003 file: standby - get metadata returned: No such file or directory
2025-06-23 14:57:43 INFO 00000 archived redo log standby is not ready for read, sleeping 10000000 us

because I use DataGuard config, so, when v$archived_log ‘s column ”name" , have tnsname for standby DB，but not redo log path！

Describe the solution you'd like
I wan’t author add arch_dest config param to define archived log list metadata filter;

[bersler]

Please describe more details about the issue. Can you please give instructions to reproduce on Docker?

[kirmorozov1992]

@bersler @zwjcool
Hello! I have the same problem. I found sql query in source code that, I think, is the source of the problem.

Log message:

2025-12-22 12:21:32 ERROR 10003 file: AMAIN_STB - get metadata returned: No such file or directory 2025-12-22 12:21:32 INFO 00000 archived redo log AMAIN_STB is not ready for read, sleeping 10000000 us

Query in code here ReplicatorOnline.h:

static constexpr std::string_view SQL_GET_ARCHIVE_LOG_LIST { "SELECT" " NAME" ", SEQUENCE#" ", FIRST_CHANGE#" ", NEXT_CHANGE#" " FROM" " SYS.V_$ARCHIVED_LOG" " WHERE" " SEQUENCE# >= :i" " AND RESETLOGS_ID = :j" " AND NAME IS NOT NULL" " ORDER BY" " SEQUENCE#" ", DEST_ID" ", IS_RECOVERY_DEST_FILE DESC" };

I execute query in database and got this result:

SELECT NAME, SEQUENCE#, FIRST_CHANGE#, NEXT_CHANGE# FROM SYS.V_$ARCHIVED_LOG WHERE SEQUENCE# >= 340062 AND RESETLOGS_ID = 920435507 AND NAME IS NOT NULL ORDER BY SEQUENCE#, DEST_ID, IS_RECOVERY_DEST_FILE DESC;

`--------------------------------------------------------------------------------
SEQUENCE# FIRST_CHANGE# NEXT_CHANGE#

---

/d7/archlog/arch__340083_1_920435507.log
340083 1.1043E+12 1.1043E+12

AMAIN_STB
340083 1.1043E+12 1.1043E+12`

[bersler]

Please describe the usage case that you would like to use and not suggestions to change particular lines in the code. I can't create any test case based on that. It really does not help with anything.

[kirmorozov1992]

@bersler

We connect to the Oracle primary node (Data Guard is enabled) and read online and archived redo logs via an sshfs mount.

While reading archived redo logs, we see the following messages in the logs:

2025-12-24 07:57:45 TRACE SQL SELECT NAME, SEQUENCE#, FIRST_CHANGE#, NEXT_CHANGE# FROM SYS.V_$ARCHIVED_LOG WHERE SEQUENCE# >= :i AND RESETLOGS_ID = :j AND NAME IS NOT NULL ORDER BY SEQUENCE#, DEST_ID, IS_RECOVERY_DEST_FILE DESC
2025-12-24 07:57:45 TRACE SQL PARAM1: 340171
2025-12-24 07:57:45 TRACE SQL PARAM2: 920435507
2025-12-24 07:57:45 TRACE REDO searching archived redo log for seq: 340171
2025-12-24 07:57:45 TRACE REDO ABC_STB is seq: 340171, scn: 1105913485623
2025-12-24 07:57:45 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory

There is another case: after an archived redo log file is fully processed, a log switch occurs and the following error appears.

Although the archived file is read completely, during the switch the connector repeatedly fails to access the next archived redo log:

2025-12-23 16:53:01 TRACE CHKPT on: 1105978134543 with switch
2025-12-23 16:53:01 TRACE PERFM 125402.439000 ms, Speed: 16.593853 MB/s, Redo log size: 2080 MB, Read size: 2080 MB, Read speed: 16.656487 MB/s, Max LWN size: 4, Supplemental redo log size: 15037034
0 bytes (6.891425 %)
2025-12-23 16:53:01 TRACE REDO ABC_STB is seq: 340172, scn: 1105952249795
2025-12-23 16:53:01 TRACE REDO ABC_STB is seq: 340173, scn: 1105978134772
2025-12-23 16:53:01 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:53:02 TRACE CHKPT write scn: 1105978134543 time: 1220643977 seq: 340173 offset: 2181992960 name: cft_src-chkpt-1105978134543
2025-12-23 16:53:11 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:53:21 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:53:31 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:53:41 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:53:51 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:54:01 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:54:11 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:54:21 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:54:31 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:54:41 ERROR 10003 file: ABC_STB - get metadata returned: No such file or directory
2025-12-23 16:54:41 ERROR 10009 file: ABC_STB - failed to open after 10 tries
2025-12-23 16:54:41 TRACE THRD stop hard

[Vesert]

I’d like to add extra information to @kirmorozov1992’s trace.
I executed the following SQL on the primary database:

SELECT DEST_ID, DEST_NAME, TARGET, DESTINATION, STATUS, ERROR, ARCHIVER, TRANSMIT_MODE
FROM v$archive_dest
WHERE valid_now = 'YES';

The query returned two destinations: local and standby (Data Guard):

DEST_ID	DEST_NAME	TARGET	DESTINATION	STATUS	ARCHIVER	TRANSMIT_MOD
1	LOG_ARCHIVE_DEST_1	PRIMARY	/xxx/local_path_on_primary	VALID	ARCH	SYNCHRONOUS
2	LOG_ARCHIVE_DEST_2	STANDBY	ABC_STB	VALID	LGWR	ASYNCHRONOUS

Next, I ran:

SELECT NAME, SEQUENCE#, FIRST_CHANGE#, NEXT_CHANGE#, APPLIED, DEST_ID
FROM SYS.V_$ARCHIVED_LOG
WHERE DEST_ID = 1;

This returned a path accessible via SSHFS.
Running the same query for DEST_ID = 2:

SELECT NAME, SEQUENCE#, FIRST_CHANGE#, NEXT_CHANGE#, APPLIED, DEST_ID
FROM SYS.V_$ARCHIVED_LOG
WHERE DEST_ID = 2;

returns a path that is not reachable via SSHFS.