ElasticGrid - create cluster may fail when user-provided VPC/SG/subnet is in use

Description

[ OK ] 9:43 Installing software
[ERROR] Failed to bootstrap Nomad

looks like nomad is trying to use a wrong interface for binding:
c2-user@ip-1-2-3-50 ~]$ systemctl status nomad
● nomad.service - nomad agent
Loaded: loaded (/etc/systemd/system/nomad.service; enabled; vendor preset: disabled)
Active: active (running) since Tue 2020-08-18 14:36:31 UTC; 59s ago
Main PID: 24511 (nomad)
Tasks: 9
Memory: 17.9M
CGroup: /system.slice/nomad.service
└─24511 /usr/bin/nomad agent -config=/etc/nomad.d/nomad.hcl -meta node_type=client
Aug 18 14:37:08 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: 2020-08-18T14:37:08.739Z [ERROR] client: error discovering nomad servers: error="3 errors occurred:
Aug 18 14:37:08 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: * rpc error: failed to get conn: dial tcp 172.17.0.1:4647: connect: connection refused
Aug 18 14:37:08 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: * rpc error: failed to get conn: dial tcp 172.17.0.1:4647: connect: connection refused
Aug 18 14:37:08 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: * rpc error: failed to get conn: dial tcp 172.17.0.1:4647: connect: connection refused
Aug 18 14:37:08 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: "
Aug 18 14:37:13 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: 2020-08-18T14:37:13.756Z [ERROR] http: request failed: method=GET path=/v1/agent/health?type=client error="{"clie..." code=500
Aug 18 14:37:13 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: 2020-08-18T14:37:13.756Z [DEBUG] http: request complete: method=GET path=/v1/agent/health?type=client duration=268.906µs
Aug 18 14:37:22 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: 2020-08-18T14:37:22.338Z [WARN] client.server_mgr: no servers available
Aug 18 14:37:23 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: 2020-08-18T14:37:23.758Z [ERROR] http: request failed: method=GET path=/v1/agent/health?type=client error="{"clie..." code=500
Aug 18 14:37:23 ip-1-2-3-50.us-east-2.compute.internal nomad[24511]: 2020-08-18T14:37:23.758Z [DEBUG] http: request complete: method=GET path=/v1/agent/health?type=client duration=178.916µs
Hint: Some lines were ellipsized, use -l to show in full


Workaround

Steps for creating a vpc + subnet + security group for GSCTL

1. Create vpc with CIDR of 172.31.0.0/16
a. Right click on the vpc -> “Edit DNS hostnames” -> enable
2. Create subnet with CIDR of 172.31.0.0/20 associated with the vpc
a. Right click on the created subnet -> “Modify auto-assign IP settings” -> enable Auto-assign IPv4
3. Create internet getaway
a. Right click -> “attach to vpc” -> to your vpc
4. Create route table associated with the vpc
a. Right click -> “Edit subnet associations” -> associate to the subnet
b. Right click -> “Edit routes” -> “add route”:
b.1. Destination = 0.0.0.0/0
b.2. Target = the internet getaway
5. Create a security group associated with the vpc, add an inbound rule for the created security group after the creation.

  • For the workaround, it is important to use the specified CIDR

Acceptance Test

n/a

Assignee

Mishel Liberman

Reporter

Max Berkman

Labels

Priority

Medium

SalesForce Case ID

None

Fix versions

Commitment Version/s

None

Due date

None

Product

None

Edition

Open Source

Platform

All
Configure