Health Monitor

Can you provide the VIP/Pool Configurations of QA & Prod as well as the status of the pools? Please whitewash IPs. list ltm virtual VIPNAME list ltm pool POOLNAME show ltm pool POOLNAME members

QA ====== virtual vip-x.x.x.x-http { snat automap destination x.x.x.x:http ip protocol tcp rules fqdn-url-redirect persist source_addr_900sec profiles { http {} tcp {} } } pool pool-x.x.x.x-55000 { monitor all sap-http-55000 members x.x.x.x:55000 {} } Ltm::Pool: pool-x.x.x.x-55000 -------------------------------------- Status Availability : available State : enabled Reason : The pool is available Traffic ServerSide Bits In 5.5M Bits Out 6.9M Packets In 2.5K Packets Out 2.1K Current Connections 0 Maximum Connections 26 Total Connections 278 Prod ========== virtual vip-x.x.x.x-https { snat automap destination x.x.x.x:https ip protocol tcp rules { fqdn-url-redirect-https Secure-Cookie } persist source_addr_1800sec profiles { sap_ssl {} tcp_timeout1800 {} fqdn_2014 { clientside } } } pool pool-x.x.x.x-54000 { monitor all tcp members { x.x.x.x:54000 {} x.x.x.x:54000 {} } } Ltm::Pool: pool-x.x.x.x-54000 -------------------------------------- Status Availability : available State : enabled Reason : The pool is available Traffic ServerSide Bits In 37.2G Bits Out 107.3G Packets In 14.5M Packets Out 17.5M Current Connections 51 Maximum Connections 533 Total Connections 1.0M

Hello, Reviewing your PROD VIP, I don't show you are doing an http monitor check on this. If this is the standard tcp monitor, even though you are replying with a 503, the pool member is going to be marked up. Also, for the QA VIP, if the pool member is still showing up even though you have set it to a 503 (and you wish it to be marked down), your http check will need to be modified.

Thanks for your response Dareuja. Currently we are using the Health Monitor only in QA, not in production since we have applied the HTTP content check in Prod URL, the first pool member went into disabled status on HTTP response (503) and the new connection is not going forward to other pool member and the connection got broken. Could you please let me know what would be the reason for this ?

Hello Antony, If this link https://devcentral.f5.com/questions/http-health-monitor-38594 is the health monitor that is applied, I may not be able to assist in troubleshooting very much. I have 0 experience using the recv disable string. Below is the SOL that goes over this option. It appears it might be possible both your recv disable and recv are matching so while the node is disabled it is still allowing existing connections. This combined with source IP persistence may be your issue. If the recv disable string is not strictly required, I would test the monitor without this. I've included a Q & A where Piotr's answer appears to shed light on your issue. Q & A: https://devcentral.f5.com/questions/session-persistnce-when-pool-disabled SOL: https://devcentral.f5.com/questions/http-health-monitor-38594

Thanks for your detailed response. I have resolved this issue by removing recv disable string and added specific alias service port.