Postgresql stream replication slow

目前有两台postgres db, 使用stream replication 做主从同步，在比较重的写操作时发生了比较重的延迟．

通过对　
pg_xlog_location_diff(pg_current_xlog_location(), sent_location) AS not_sent,
pg_xlog_location_diff(sent_location, write_location) AS not_written,
pg_xlog_location_diff(write_location, flush_location) AS not_flushed,
pg_xlog_location_diff(flush_location, replay_location) AS not_replayed
FROM pg_stat_replication;
这几个参数的监控，我们发现在
pg_current_xlog_location() - (sent_location)有比较重的延迟，延迟的size将近在100MB+ ，网络吞吐量最大值在１０MB/s.
最大的延迟将近５分钟．

本地模拟：
更新数据表的字段　一个表将近110w数据　，size 是 612 MB.
我们在观察主从两边的wal　log 的数目时发现，此次操作总归会生产104个wal 文件　size =16MB* 104 =1.644268775GB．
手工对比主从库wal　文件状态，从库远远落后于主库，将近落后３０多个wal file.

我也使用了iperf去测试了网络情况

Client connecting to 192.168.6.55, TCP port 5001

TCP window size: 125 KByte (default)

[ 3] local 192.168.6.56 port 41489 connected with 192.168.6.55 port 5001
[ ID] Interval Transfer Bandwidth
[ 3] 0.0-10.0 sec 1.09 GBytes 939 Mbits/sec

网卡最大值为1G.

数据库 MAX_SEND_SIZE (XLOG_BLCKSZ * 16)
XLOG_BLCKSZ= 8K

理论上　延迟最大值为一秒．

Postgres version:psql (PostgreSQL) 9.2.13

有没有其他参数会影响到stream replication的性能？

期待大家的回复，讨论．