Network switch or routing issue?

Hi Guys,

Seems like I'm having this weird issue on my network that a specific IP block under switch1 vrf cust instance causing some delay or issue when accessing or connecting to a public proxy server and sometimes this cause delay when browsing using the said proxy server.

Topology: https://ibb.co/LCVc2bC

Attached the topology to further understand this issue, Currently we have 2 physical switch both are in stack (in pair). Switch1 pair of 3850 while switch 2 pair of 3750... To make the story short, We already run an isolation and prove the below when accessing/connecting to public proxy.

1. When Segment A/B is forwarding from SW1(vrf cust) -> sw1 global -> R1(private network/dc internet) both segment can connect to proxy and browse to internet without any issue.
2. When VIP under switch2 vrf cust SW2(vrf cust) -> sw2 global -> R2(internet router) segment can connect to proxy and browse to internet without any issue.
3. When segment VIP is under switch1 vrf cust then issue occurred.

From the topology you can see that a po1 is being used by sw1 and sw2 for L2 and layer3 interconnection. From the isolation we've conduct it is showing the issue take placed during the layer3 forwarding from sw1-sw2 (global) via EIGRP. although we can see that L2 forwarding to reach VIP @ sw2 vrf cust is being used by segment B but without any issue.

​

Checked the ff. on Sw1-Sw2 port-channel.

1. No error or drop seen on physical interface of PO member ports.
2. EIGRP is stable, and since access to private network doesn't have any not sure if any issue on switch global for web or specific ports.
3. Optical value is good and ping test between eigrp (SVI) doesn't have any issue.
4. no recent logs or congestion on any of the links.
5. tried shutting down 1 physical member port of PO to test if any issue with loadbalancing but issue still persist.
6. no configuration changes.
7. no drops on qos applied on PO and no drop at Copp.
8. Switch utilization are all normal.
9. routing is correct. and sw stacking etc.

​

Question:

1. What could go wrong with this setup? Why I'm having issue with specific destination(proxy server) traversing to sw1-sw2 (global) eigrp link? While other traffic passing thru this link is working?
2. What other things should be check?

​

anyone encountered this kind of issue with the said platform?

Thank you