AWS AMI performance issue?

[mofoster]

Hi,

I've recently encounter some performance issue when I deploy AMI version of VestaCP 0.9.8-17 on Ubuntu 16

What I've done was setup Wordpress + Woocommerce on VestaCP 0.9.8-17 Ubuntu 16 in Region A, take AMI snapshot, then deploy to Region C, and run an ELB load balancer on top of it. So A is the original copy of the instance, and C is exact the same as A. I did change the host name and that's about it. and the local IP is served automatically from VPC DHCP server.

However do noticed that C is about 50% slower than A (1000ms vs 500ms), i am wondering if there's any other configuration change needs to be make after I launch C?

Thanks in advance :)

[skurudo]

C is about 50% slower than A

And why so? (if servers specs looks identycal, right?)
top/htop say something? Problem with cpu/ram/IO ?

[mofoster]

C is about 50% slower than A

And why so? (if servers specs looks identycal, right?)
top/htop say something? Problem with cpu/ram/IO ?

Yeah, server spec is exactly the same, nothing with processing / resource, so I am suspected something need to be re-config after I launch the instance, e.g. something similar to v-update-sys-ip?

[tjebbeke]

Network Latency or distance. Also you are using a load balancer? So your connection is going through the load balancer this has also it's latency.

[mofoster]

Network Latency or distance. Also you are using a load balancer? So your connection is going through the load balancer this has also it's latency.

Yes it's going through LB, however if it's that then both service should have the same latency issue...

[skurudo]

LB - one latency - region A
LB - other latency - region C

One LB, two different latency, no?
Did you try ask some from AWS support about this?

[mofoster]

It turn out to be the network latency between web and other services (redis cache & db), web is in a different AZ than the two.