今天到客户的生产库查看alter日志，发现反复的报如下的错误：

Thread 1 cannot allocate new log, sequence 36676
Private strand flush not complete

初步判断为redo太小了切换太频繁，buffer中的dirty block没有完全些人数据文件的时候，日志又切换了，造成无redo log可用。

首先查看了redo的情况

SQL> select group#,thread#,bytes/1024/1024||'MB' from v$log;

 

    GROUP#    THREAD# BYTES/1024/1024||'MB'

---------- ---------- ------------------------------------------

         1          1 50MB

         2          1 50MB

         3          1 50MB

         4          1 50MB

         5          1 50MB

SQL> select group#,member from v$logfile;

 

 

    GROUP# MEMBER

---------- ---------------------------------------------

         1 /oradata/LEGATO/legato_redo01a.rdo

         1 /oradata/LEGATO/legato_redo01b.rdo

         2 /oradata/LEGATO/legato_redo02a.rdo

         2 /oradata/LEGATO/legato_redo02b.rdo

         3 /oradata/LEGATO/legato_redo03a.rdo

         3 /oradata/LEGATO/legato_redo03b.rdo

         4 /oradata/LEGATO/legato_redo04a.rdo

         4 /oradata/LEGATO/legato_redo04b.rdo

         5 /oradata/LEGATO/legato_redo05a.rdo

         5 /oradata/LEGATO/legato_redo05b.rdo

50MB确实是有点小，在alter中查看大约三分钟就切换一次日志，手动调整了一下redo


SQL> select group#,bytes/1024/1024||'MB',status from v$log;

    GROUP# BYTES/1024/1024||'MB'                      STATUS
---------- ------------------------------------------ --------------------
         1 300MB                                      CURRENT
         2 300MB                                      INACTIVE
         3 300MB                                      INACTIVE
         4 300MB                                      INACTIVE
         5 300MB                                      INACTIVE
         6 300MB                                      INACTIVE
         7 300MB                                      INACTIVE
         8 300MB                                      INACTIVE
         9 300MB                                      INACTIVE
        10 300MB                                      INACTIVE

到目前为止还没有再次出现题目中的报错。