- Newest
- Most votes
- Most comments
Hi,
I think that you really want to setup Infrastructure Performance from AWS Network Manager in your case to obtain the cross-region latency metrics that you are looking for. See https://docs.aws.amazon.com/network-manager/latest/infrastructure-performance/what-is-nmip.html
Infrastructure Performance allows you to obtain near real-time and historical network
latency across AWS Regions and across or within Availability Zones for a specified time period.
All details in https://docs.aws.amazon.com/network-manager/latest/infrastructure-performance/how-nmip-works.html:
Network latency measurements are an aggregate of latency performance information
captured by active probing between AWS managed probes within the AWS global network.
Inter-Region latency metrics are generated by aggregating latency measurements from
probes located across AWS Regions, filtered on your chosen source and destination AWS Region pairs.
Inter-Availability Zone latency metrics are generated by aggregating latency measurements
from probes located across Availability Zones, filtered on your chosen source and destination
Availability Zones.
Intra-Availability Zone latency metrics are generated by aggregating latency measurements
between all probes within a single Availability Zone. This includes all probes that are across
AWS data centers, and within the same data center, for the chosen Availability Zone.
The placement of the probes is not related to EC2 instances or AWS services running in your
account. For these reasons, the same latency metric data is presented in all AWS accounts.
Best,
Didier
Do you mean that the TGWs in both the regions are attached to the same DXGW? Or that there are two DXGWs with transit VIFs leading to different on-premises locations that are connected to one another?
The normal and recommended way to interconnect two TGWs would be to use transit gateway peering (https://docs.aws.amazon.com/vpc/latest/tgw/tgw-peering.html). Are you intentionally avoiding establishing a peering connection?
Relevant content
asked 4 months ago
asked 4 years ago
asked 3 years ago
- AWS OFFICIALUpdated 10 months ago
Making measurements certainly answers the immediate question of whether there's unexpectedly high latency, but if there are on-premises elements outside AWS technologies involved, I'd advise strong caution assuming that the connection quality metrics are going to remain that way. I'd heavily recommend using transit gateway peering as a solution AWS manages end to end, if there's nothing preventing doing so.