Upgrade T2 EC2 instances with T3 ones

[tzouris]

Update T2 EC2 instances to the newer, faster and cheaper T3. Compared to the T2 type, T3 instances are more energy efficient and have an overall lower environmental impact.

[tzouris]

Upgraded:

openvpnaccess From t2.micro to t3.micro t2.micro: $0.0116 per hour, 1 vCPU, 1 GiB RAM, Low to Moderate network performance t3.micro: $0.0104 per hour, 2 vCPUs, 1 GiB RAM, Up to 5 Gigabit

invenio-dev From t2.large to t3.large t2.large: $0.0928 per hour, 2 vCPUs, 8 GiB RAM, Low to Moderate network performance t3.large: $0.0832 per hour, 2 vCPUs, 8 GiB RAM, Up to 5 Gigabit

[tzouris]

Successfully upgraded:

From t2.medium to t3.medium:

t2.medium: $0.0464 per hour, 2 vCPUs, 4 GiB mem, Low to Moderate network performance t3.medium: $0.0416 per hour, 2 vCPUs, 4 GiB mem, Up to 5 Gigabit

hcommons-staging hcommons-dev-wordpress hcommons-dev-2 hcommons-dev-3

From t2.small to t3.small:

t2.small: $0.023per hour, 1 vCPU, 2 GiB mem, Low to Moderate network performance t3.small: $0.0208 per hour, 2 vCPUs, 2 GiB mem, Up to 5 Gigabit

hcommons-dev-proxy-0

The following instances can't be upgraded because they require enhanced networking with the Elastic Network Adapter (ENA).

hcommons-dev-hc-idp hcommons-dev-registry This is because they were launched from an old AMI, that did not support ENA at the time. There is no point in dealing with updating these two dev instances at the moment only to upgrade them to t3 type since they will be deprecated when we set up new EC2 IDMS instances and eventually move to local development.

[tzouris]

Production instances will be upgraded too, but we'll send a notice to our users first letting them know there will be a short service interruption.

[tzouris]

The only t-type production instance that can be upgraded right now is proxy. I suppose that both hc-idp and registry will have the same ENA issue.