Issues with deployment
# support
s
Hello all, I'm new here 🙂 I'm trying to deploy medplum on AWS and I'm struggling with the following issue while executing the synth step : [Error at /MedplumStaging/BackEnd] Found zones: [] for dns:mindbowser.com, privateZone:undefined, vpcId:undefined, but wanted exactly 1 zone
r
Hi @sanket_89360 , I assume you're using Medplum for your SSL certs. This error usually means that you don't have the corresponding "Hosted Zones" inside your Route53 account. This is a manual step that needs to be done before running our automation scripts. Were you able to follow the instructions here (https://www.medplum.com/docs/self-hosting/install-on-aws#configure-your-name-servers), under "Configuring your nameservers"
s
but i have already created the zones in route53
r
Could you share more details about the zone that has been set up? This error message is coming directly from AWS saying that it can't find zones from mindbowser.com
Copy code
[Error at /MedplumStaging/BackEnd] Found zones: [] for dns:mindbowser.com, privateZone:undefined, vpcId:undefined, but wanted exactly 1 zone
s
aws route53 list-hosted-zones { "HostedZones": [ { "Id": "/hostedzone/Z0394455PI5IKC6SKSPF", "Name": "healthconnect.mindbowser.com.", "CallerReference": "2dde0433-7dad-4a9f-b68a-3b27f6025332", "Config": { "Comment": "", "PrivateZone": false }, "ResourceRecordSetCount": 3 } ] }
r
@sanket_89360 , looking at this code, we are looking for a zone for the domain
mindbowser.com
. Do you have a zone for just this domain somewhere?
s
domain "mindbowser.com" is registered on go-daddy and on aws route53 i have created zone as "healthconnect.mindbowser.com".
r
You will need to configure "custom name servers" on go daddy for mindbowser.com. See the link to the AWS docs in this section of the medplum documention for instructions: https://www.medplum.com/docs/self-hosting/install-on-aws#configure-your-name-servers
s
"mindbowser.com" is main domain which is in use. I want to use subdomain "healthconnect.mindbowser.com", can you guide me how to achieve this?
r
You will have to configure custom name servers for the subdomain
healthconnect.mindbowser.com"
in go daddy https://www.godaddy.com/help/add-an-ns-record-19212
s
i have already performed above step, still getting same error. error: [Error at /MedplumStaging/BackEnd] Found zones: [] for dns:mindbowser.com, privateZone:undefined, vpcId:undefined, but wanted exactly 1 zone
r
hi @sanket_89360 , we are updating our aws tooling to allow you to use a hosted zone at
healthconnect.mindbowser.com"
Currently, our automated scripts only work with 2-part hosted zone names. I will keep you updated with the resolution
If you need a workaround, you can try to configure SSL certificates for the following URLS outside of the medplum tooling
app.healthconnect.mindbowser.com
api.healthconnect.mindbowser.com
storage.healthconnect.mindbowser.com
s
i am using already created certificate "*.mindbowser.com"
p
@rahul1 any thoughts here?
Any other way to fix for time being so we can continue with the deployment?
r
@Pravin , have you considered using the
skipDns
configuration to bypass route53 / SSL setup? it sounds like maybe you've done this manually already? https://www.medplum.com/docs/self-hosting/config-settings
p
@sanket_89360 please check above option and see
r
I can provide a bit more background on the error message you were getting. - The error message you got comes from thsi part of our deploy scripts (https://github.com/medplum/medplum/blob/7180d87a0d3b38d7820a54f8937f89d286be19c2/packages/cdk/src/backend.ts#L423-L434) - What it is trying to do is to automate the creation of the A record to point from your API domain name to the ELB load balancer that was set up to host the server - However, if you are ok doing all of your DNS configuration manually, the
skipDns
option will skip over this step on the backend setup automation
s
As per your suggestion, I proceeded with the skipDns option. Now, I am stuck at the deploy stage. The issue is related to the creation of the CloudFormation stack. It is failing during the creation of the backend Fargate service. Can you help me to resolve this issue?
r
hi @sanket_89360 , could you create a new support ticket with a description of the error you are seeing. Putting as much information as possible will help me troubleshoot for you
p
@sanket_89360 please also note down different issues you faced and solved, can be useful for troubleshooting or add in medium deployment doc for new users
s
ok
c
hi @rahul1 , any update on allowing hosted zones on subdomains?
r
@casualgardener - have you tried using the
hostedZoneName
configuration setting for the server configs? https://www.medplum.com/docs/self-hosting/config-settings
c
After I fixed my nameserver config, I got past this error. Then I had to manually "Create records in Route 53" within Certificate Manager (per step 1 of the second answer here: https://stackoverflow.com/questions/40761309/adding-ssl-to-domain-hosted-on-route-53-aws) on the pending certificates to the subdomain (e.g. api.sub.domainname.com) the bootstrap command created, to fix SSL errors I ran into when deploying.
Maybe hostedZoneName would have fixed though - I didn't try it.
(and just to be clear, not implying these are MedPlum errors - just a new user with not-much experience deploying on AWS)
r
Thanks for circling back @casualgardener ! The community will benefit from your experience
177 Views