Deployment stuck on "Waiting until master instances are elected"

[dmitryyankowski (John Smith)]

Hi guys, i'm trying to deploy another "allocator" instance to my deployment.. and it's currently stuck on "Waiting until master instances are elected". What am I doing wrong here. Is deploying 2 instances not allowed because of quorum rules or something? Is there some other configuration I should be adding to the deployment to specify which instance is my master?

The command I used to add the new allocator was:
bash <(curl -fsSL https://download.elastic.co/cloud/elastic-cloud-enterprise.sh) install --coordinator-host IP_ADDRESS --roles-token TOKEN --availability-zone ece-zone-2 --roles "allocator"

Any help would be appreciated!

Also seeing this error!

There was a problem applying this configuration change!

Plan change failed: [ClusterFailure:MasterNotDetectedInTime]: Failed to detect that a master was elected. Check thee health of the cluster, and look at the instance logs to determine if there were any issues with cluster formation or leadership election. Details: [{"details":"The state did not become the desired one before [5 minutes] elapsed. Last error was: [Instance [instance-0000000017] has no master.]"}]

[dmitryyankowski (John Smith)]

I've run into this issue on multiple different cloud hosts now. Tried with bare metal servers I operate locally and on AWS EC2. There's got to be something i'm missing here.

According to High availability | Elastic Cloud Enterprise Reference [2.8] | Elastic it says:

The main difference between Elastic Cloud Enterprise installations that include two or three availability zones is that three availability zones enable Elastic Cloud Enterprise to create clusters with a tiebreaker .

However when I try to make this change:

On my activity feed it just says: Waiting until master instances are elected for like 5 minutes, and then ends up failing with the following error:

Plan change failed: [ClusterFailure:MasterNotDetectedInTime]: Failed to detect that a master was elected. Check thee health of the cluster, and look at the instance logs to determine if there were any issues with cluster formation or leadership election. Details: [{"details":"The state did not become the desired one before [5 minutes] elapsed. Last error was: [Instance [instance-0000000017] has no master.]"}]

Is there a step im missing somewhere? I figured deploying Elasticsearch to 2 zones would be pretty simple.. as everything else is. It doesn't create a Tiebreaker node in the 3rd zone.. like it does in Elastic Cloud (Non self hosted).

Note: This is my current setup:

I'm assuming other people have run into this as well... since it's not just tied to 1 cloud provider.

[dmitryyankowski (John Smith)]

Look's like the issue was with the security group inside EC2. One of my EC2 instances didn't have the inbound ports open. Derp me! Works just fine now!