一位QQ上的网友问我的一个问题，我觉得比较有意思。
记录如下 :
Question :
我在启动POSTGRES的STANDBY数据库时，报
--------------------------
LOG: database system was shut down in recovery at 2011-12-30 23:20:25 CST
LOG: entering standby mode
LOG: restored log file "0000000100000002000000E9" from archive
LOG: invalid record length at 2/E9000108
LOG: invalid record length at 2/E9000108
LOG: streaming replication successfully connected to primary
LOG: invalid record length at 2/E9000108
FATAL: terminating walreceiver process due to administrator command
LOG: restored log file "0000000100000002000000E9" from archive
LOG: invalid record length at 2/E9000108
FATAL: the database system is starting up
LOG: invalid record length at 2/E9000108
LOG: restored log file "0000000100000002000000E9" from archive
LOG: invalid record length at 2/E9000108
--------------------------------------
我是把主库整个做了个TAR 包到备库的
问题分析 :
报错invalid record length at 2/E9000108
源码中对应的部分如下 :
/*
* NOTE: We disallow len == 0 because it provides a useful bit of extra
* error checking in ReadRecord. This means that all callers of
* XLogInsert must supply at least some not-in-a-buffer data. However, we
* make an exception for XLOG SWITCH records because we don't want them to
* ever cross a segment boundary.
*/
if (len == 0 && !isLogSwitch)
elog(PANIC, "invalid xlog record length %u", len);
START_CRIT_SECTION();
因此报错的原因是standby在恢复过程中从 xlog 0000000100000002000000E9 的 2/E9000108位置读取到了len == 0 并且 !isLogSwitch 的信息。
日志中还提到了restored log file "0000000100000002000000E9" from archive
因此问他要了recovery.conf 配置文件的内容如下 :
standby_mode = 'on'
primary_conninfo = 'host=隐藏 port=5432 user=隐藏 password=隐藏'
restore_command = 'cp "/home/postgres/data/pg_xlog/%f" "%p"'
trigger_file = '/home/postgres/trigger_activestb'
从这个配置文件来看，restore_command = 'cp "/home/postgres/data/pg_xlog/%f" "%p"' 有问题，如果要配置restore_command，那么应该是把主库的归档文件拷贝standby的pg_xlog目录。
要么就是不要配置restore_command , 那么standby会尝试从stream获取xlog的信息（要求primary库的这个xlog没有被覆盖）。
Answer :
经过确认主库的 0000000100000002000000E9 存在pg_xlog中未被覆盖.
注释standby节点recovery.conf的restore_command = 'cp "/home/postgres/data/pg_xlog/%f" "%p"'
重启standby.
未解决，还是报 invalid record length at 2/E9000108
另外需要了解的是，
做STANDBY的时候有没有按照正常流程来做( select pg_start_backup(), backup, select pg_stop_backup() )
拷贝数据文件的过程中有没有错误发生.
这些都不得而知, 因此问题不太好判断。
所以先重做 standby 来解决目前的问题 .
最后.
可以尝试使用pg_filedump来分析一下 0000000100000002000000E9的内容。（由于时间关系，暂时没做）
【参考】
src/backend/access/transam/xlog.c
http://blog.163.com/digoal@126/blog/static/163877040201142610215685/