轉 Alert.log shows No Standby Redo Logfiles Of Size 153600 Blocks Available

http://blog.itpub.net/23135684/viewspace-703620/

 

Alert.log shows No Standby Redo Logfiles Of Size 153600 Blocks Available [ID 405836.1]
		修改時間 05-APR-2011     類型 PROBLEM     狀態 PUBLISHED

In this Document
  Symptoms
  Cause
  Solution

---

 

 

Applies to:

Oracle Server - Enterprise Edition - Version: 10.1.0.2 to 10.2.0.5 - Release: 10.1 to 10.2
Information in this document applies to any platform.
***Checked for relevance on 05-APR-2011***

Symptoms

 After real time apply works within 24hrs
the following error occurs.

ERROR
Rfs[2]: no standby redo logfiles of size 153600 blocks available

Checking the standby logs on standby database, all SRLs are ACTIVE dated from weeks ago- normally we see 1 used for 1 thread and the others will be UNASSIGNED

standby> select * from v$standby_log;
GROUP# DBID THREAD# SEQUENCE# BYTES USED ARC STATUS FIRST_CHANGE# FIRST_TIME LAST_CHANGE# LAST_TIME
---------- ---------------------------------------- ---------- ---------- ---------- ---------- --- ---------- ------------- -------------------- ------------ --------------------
7 2924799764 1 233641 157286400 148177920 NO ACTIVE 8.2227E+12 01-Dec-2006 21:36:20 8.2227E+12 01-Dec-2006 21:54:04
8 2924799764 1 233637 157286400 148179968 NO ACTIVE 8.2227E+12 01-Dec-2006 20:20:21 8.2227E+12 01-Dec-2006 20:39:22
9 2924799764 1 233635 157286400 148292608 NO ACTIVE 8.2227E+12 01-Dec-2006 19:43:15 8.2227E+12 01-Dec-2006 20:00:54
10 2924799764 1 233627 157286400 12578816 NO ACTIVE 8.2227E+12 01-Dec-2006 18:00:25 8.2227E+12 01-Dec-2006 18:01:31
11 2924799764 1 233633 157286400 148156416 NO ACTIVE 8.2227E+12 01-Dec-2006 19:06:06 8.2227E+12 01-Dec-2006 19:22:41
12 2924799764 1 233639 157286400 148199424 NO ACTIVE 8.2227E+12 01-Dec-2006 20:57:35 8.2227E+12 01-Dec-2006 21:16:32
13 2924799764 1 233629 157286400 148305920 NO ACTIVE 8.2227E+12 01-Dec-2006 18:01:41 8.2227E+12 01-Dec-2006 18:20:14
14 2924799764 1 233631 157286400 148451328 NO ACTIVE 8.2227E+12 01-Dec-2006 18:32:17 8.2227E+12 01-Dec-2006 18:48:59
15 2924799764 1 234647 157286400 3922944 YES ACTIVE 8.2228E+12 20-Dec-2006 16:37:31 8.2228E+12 20-Dec-2006 16:37:57

Cause

In Dec1 SRLs created on LG were not archived/stuck and hence remained ACTIVE and could not longer be assigned. At the that time we see that Primary was archiving every minute and only 1 ARCH available to archive to standby. From standby log_archive_max_processes was set to 2, hence only 1 ARCH archiving Locally and most likely unable to cope with the amount of archiving required.

Solution

1. On Standby/Primary set log_archive_max_processes=10

alter system set log_archive_max_processes=10 scope=both;

OR

2. Another way around this if the logs have been applied as they have in this case, the old standby logs 
can be dropped and recreated to clear the problem.

alter database drop standby logfile '<logfile_name>';
then
alter database add standby logfile group x '<logfile_name>';

<logfile_name><logfile_name>

 

#########################################################

 

報錯 ：

 

Sun Oct 21 04:39:06 CST 2018

 

FAL[client]: Failed to request gap sequence

 

GAP - thread 2 sequence 101986-101989

 

 

 

 

 

解決方法1:

 

 

 

ALTER DATABASE REGISTER LOGFILE '/tmp/dba/1/o1_mf_2_101986_fwps5fbo_.arc';

 

 

 

ALTER DATABASE REGISTER LOGFILE '/tmp/dba/1/o1_mf_2_101988_fwps7ntp_.arc';

 

 

 

ALTER DATABASE REGISTER LOGFILE '/tmp/dba/1/o1_mf_2_101987_fwps7o6j_.arc';

 

 

 

ALTER DATABASE REGISTER LOGFILE '/tmp/dba/1/o1_mf_2_101989_fwpsg14g_.arc';

 

 

 

ALTER DATABASE REGISTER LOGFILE '/tmp/dba/1/o1_mf_2_102004_fwq5wq6m_.arc';

 

ALTER DATABASE REGISTER LOGFILE '/tmp/dba/1/o1_mf_2_102005_fwq5wx6s_.arc';

 

解決方法2：

Cause 3 : ARC process on the primary that is responsible for the heartbeat sticks on the network due to bug 6113783

You could see different symptoms. For example, you get fal archive error as the primary arch process sticks and it doesn't ship the missing archive log files to the standby. The worst case you see is all arch processes stick and no one does local archiving and all online redo log files are full and that causes the primary database hangs. Please see the following errors related to this cause.

Mon Oct 22 13:30:44 CST 2018                                                   <- RFS fetch gap begin failed 

FAL[client]: Failed to request gap sequence                           
GAP - thread 1 sequence 100641-100641
DBID 2583306402 branch 777418146
FAL[client]: All defined FAL servers have been attempted.
-------------------------------------------------------------
Check that the CONTROL_FILE_RECORD_KEEP_TIME initialization
parameter is defined to a value that is sufficiently large
enough to maintain adequate log switch information to resolve
archivelog gaps.
-------------------------------------------------------------
Mon Oct 22 13:32:44 CST 2018                                                               <- find the reason is  network disconnet 
RFS[342]: Possible network disconnect with primary database
Mon Oct 22 13:32:46 CST 2018
RFS[341]: Possible network disconnect with primary database
Mon Oct 22 13:52:16 CST 2018                                                               <- after 30 minutes ,RFS  retry .
Redo Shipping Client Connected as PUBLIC
-- Connected User is Valid
RFS[350]: Assigned to RFS process 23920798

 

 

FAL[client]: Failed to request gap sequence 
GAP - thread 1 sequence 2262-2266 
DBID 2591313681 branch 686760021 
FAL[client]: All defined FAL servers have been attempted.

SQL> alter session set nls_date_format='YYYY-MON-DD HH24:MI:SS'; 
SQL> select message, timestamp 
from v$dataguard_status 
where severity in ('Error','Fatal') 
order by timestamp; 

MESSAGE 
------------------------------------------------------------------------------- 
TIMESTAMP 
-------------------------- 
Error 12154 received logging on to the standby 
OCT-12-2010 15:55:45 

PING[ARC1]: Heartbeat failed to connect to standby 'STDBY'. Error is 12154. 
OCT-12-2010 15:55:45

Solution 3: The bug 6113783 was fixed in 11.2.0.2. You could workaround the issue by killing the arch processes on the primary database.

in primary :

This won't harm your primary database at all as arch processes will be respawned automatically immediately by Oracle.

    % ps -ef |grep -i arc 
    % kill -9 <ospid of arc process> <ospid of arc process> <ospid of arc process>...

   % ps -ef |grep -i arc 

   (new arch process will respawned quickly,monitor)

NOTE: If you are on Windows Platform you can use either 'orakill'-Command or any OS-Tool to kill Windows Processes/Threads

sqlplus / as sysdba

alter system switch logfile;

 

in standby db:

just cancle mrp and restart standby instance.