source address persistence failed

in a very short period of time

If source persistency is configured correctly, you should be able to reach the same backend server within 180 sec (default) after connection closure. Can you specify timings accurately?

1800 sec

What virtual server type are you using? Are you using OneConnect?

could you post the following output (remove what is not relevant):

list /ltm virtual list /ltm pool list /ltm node

hi Gonzalex: vs:L4 , i do not use oneconnect.

virtual pic.btr.lenovo.com.cn_124.127.169.133 { pool pic.btr.lenovo.com.cn destination 124.127.169.133:any persist source_addr_600 } pool pic.btr.lenovo.com.cn { lb method member least conn monitor all gateway_icmp members { 10.96.173.29:any {} 10.96.173.30:any {} } }

the system version is 10.2.4 HF 7

Ok ... but persistence will not necessary connect you always to the same backend server. Persistence will keep the same connection to the same server. But subsequent connections would still be load-balanced.

I don't spot anything wrong with your config. maybe others would.

Hi, please could you show your configuration for the persistence profile source_addr_600 and verify if a persistence record is created when accessing the VS.

tmsh show ltm persistence persist-records client-addr 

(other switches are available key, mode, node-addr, node-port, pool, virtual )

I would suggest running tcpdump for your connection to the VIP. Also, why do you need to use source ip as persistence. I have dealt with such scenario and switched to cookie persistence and the connectivity and persistence worked ok for me. The method of persistence does not only depend upon your discretion but also the application requirement.