Hello, Yes, that was my worry but I don't see how the code can fail, so it looks fine to me. Regards -- Julian Anastasov <ja@xxxxxx> -- To unsubscribe from this list: send the line "unsubscribe lvs-d
Hi, Hello, I guess, we here try to avoid old sync message to expire new connection. The problem is that UDP conns and templates are always IP_VS_CONN_F_INACTIVE, may be a TCP+SCTP protocol check is n
Hello, OK, just make sure patch has short lines (80): I assume we will not stop here sync for some connection that was normally expired in master but was delayed in backup. TCP sync starts for EST st
Hi, Hello, What if we make the backup server purge the old entry, just like the active server does, and ignore old updates if needed? Please note the new hunk on ip_vs_sync.c. This is the patch I'm u
Hello, Good idea If we want SCTP support may be we have to check our IP_VS_TCP_S_TIME_WAIT state. But cp->state checks should be per protocol, eg: /* Controlled (FTP DATA or persistence)? */ if (cp->
Hi Julian, What if we make the backup server purge the old entry, just like the active server does, and ignore old updates if needed? Please note the new hunk on ip_vs_sync.c. This is the patch I'm u
Hello, I think, the cp->n_control check is not needed, we can live with many connections... in master server. But it looks like the new mechanism opens the door for problems with the sync protocol. T
Hello, Yes, for IPVS-NAT TIME_WAIT means both sides closed. Agreed, a sysctl var is ok. Only that my preference is to have sysctl var that is checked even before is_new_conn() call to save some CPU c
Hello, Signed-off-by: Marcelo Ricardo Leitner <mleitner@xxxxxxxxxx> -- Notes: Hi, We have a report that not doing so may cause poor load balacing if applications reuse src port. With a patch like thi
Hello, Signed-off-by: Marcelo Ricardo Leitner <mleitner@xxxxxxxxxx> -- Notes: Hi, We have a report that not doing so may cause poor load balacing if applications reuse src port. With a patch like thi
Hello, People complained about UDP, such as RADIUS, etc. I guess, for TCP it is some local client that can benefit from balancing, it can be also a local testing tool. I think, checking of SEQ will n
(I'll repost if the idea is accepted, this one is just for discussion) Notes: Hi, We have a report that not doing so may cause poor load balacing if applications reuse src port. With a patch like thi
Signed-off-by: Marcelo Ricardo Leitner <mleitner@xxxxxxxxxx> -- Notes: Hi, We have a report that not doing so may cause poor load balacing if applications reuse src port. With a patch like this, it w