@bitter-dentist-28132 For anyone else having this issue, there is some discussion about it in the DO forums. It appears to be a known issue in DO, the workaround is to refresh your kubeconfig manually using doctl. Anyone else know anything about this?

only that i've had to do this. is there a way to

exec

only after a resource has been created? like the only way i can think of to have pulumi work with DO, is to do a

doctl k8s cluster kubeconfig show

on every pulumi run, and use its output as the provider data. but that also means that a run that creates the cluster, has to wait for the cluster creation to finish before running

doctl

.

are arbitrary pulumi resources thenable? in which case i guess i could just make the system call inside a

.then

@bitter-dentist-28132 There is a bug for this in the TerraformProvider https://github.com/terraform-providers/terraform-provider-digitalocean/issues/234

Hi. How does register events that should fire on

up

or

destroy

events? https://stackoverflow.com/q/57715935/14731

Specifically, I want to interact with an ec2 server after it's up (copy files to it, spawn new processes etc) and it's not clear what is the right way to do this. Currently I dumped my code into the main code block and it's running twice (which is bad). Once for the "diagnostics" step and once for "update" step. Any ideas?

https://www.pulumi.com/docs/intro/concepts/programming-model/#diff-id-olds-news states that "olds" is the old output values and "news" is the new input values. Is this a typo? Shouldn't they both be inputs or outputs?

Is there a way for a dynamic resource to mirror the lifecycle of another resource? I want to create my custom resource after an e2c instance gets created. I want to destroy it before that e2c instance gets created. Right now they live alongside each other but there is no real linking between the two.

depends

will probably help me create my dynamic resource after the e2c instance is created, but does

depends

guarantee destruction in the reverse order?

does anyone here install cert-manager with a clusterissuer with pulumi? i always get a failure on the first run:

kubernetes:<http://certmanager.k8s.io:ClusterIssuer|certmanager.k8s.io:ClusterIssuer> (letsencrypt-prod):
    error: Plan apply failed: resource letsencrypt-prod was not successfully created by the Kubernetes API server : Internal error occurred: failed calling webhook "<http://clusterissuers.admission.certmanager.k8s.io|clusterissuers.admission.certmanager.k8s.io>": the server is currently unable to handle the request

subsequent `up`s work

I'm trying to use the new relic pulumi provider and trying to use

InfraAlertCondition

https://github.com/pulumi/pulumi-newrelic/blob/master/sdk/nodejs/infraAlertCondition.ts This is not working for me This link provides all the data available through NR: https://docs.newrelic.com/docs/integrations/kubernetes-integration/understand-use-data/understand-use-data

Is there a way to

pulumi refresh

specific resources?

that'd be very useful

Also another question - can I get a region from an aws provider? Or maybe I can create an AWS client from an aws provider, so that I’ll use the AWS SDK with the corresponding configuration

Hi guys. I created a dynamic resource that attempts to ssh into an ec2 instance. I depend on

Instance.publicIp

but I am getting a connection refused trying to ssh in. If I add a short delay I can connect just fine. My guess is that depending on

publicIp

is not enough. What should I be depending on to make sure that port 22 is open and sshd is listening?

More specifically, I am getting

error: Plan apply failed: connect ECONNREFUSED 54.183.130.128:22

Hi! I'm trying to use the PostgreSQL module along with creating the PostgreSQL db itself and I'm getting a bit of an ordering problem, here's my script: https://gist.github.com/anaisbetts/e59b9af4c2300043cfb711a058e63afc

It seems as if the PostgreSQL module tries to connect immediately to the database as the script runs

Instead of when I want it to, during the actual build execution (i.e. after the database has been created)

Hi, I'm trying to follow the development of different pulumi packages, but I feel it's not an easy thing to do. You have changelogs for every repo with the necessary information, but as far as I know, it's not possible to watch 1 specific file for changes. I'm trying to use the "Releases only" watch future, but it does not always work, because if I'm not mistaken, you not always using releases, just tagging a specific commit, which is not triggering the watch notification. For example check the kubernetes repo: https://github.com/pulumi/pulumi-kubernetes/releases I got notification about the v0.25.6 release, but there are 3 new tags, but I didn't get any notification about those. Other repos are worse, e.g. the pulumi/pulumi repo does not have a single release, just tags. You have an announcements channel and you post the major updates, but not all. Would it be possible to create a new release for every single version in every repo? No need to put any description in it, just to trigger the notification. You can even create pre-releases for versions like the 1.0.0-beta2 or the new rc releases. I know it's not a big deal, but with that many repositories and fast development/lots of version could be annoying.

app.pulumi.com, dropdown where one usually selects organization doesn't work. going directly to org page continue to work though

hm, whats the motivation for kubernetes .tranformations (https://www.pulumi.com/docs/reference/pkg/nodejs/pulumi/kubernetes/yaml/#ConfigFileOpts-transformations) to return

void[]

, not void?

is there a way to wait for things, like what you can do with terraform and

null_resource

? I just can't seem to order a

ClusterIssuer

correctly so i'd like to just wait a bit after cert-manager is installed.

Plan apply failed: Unauthorized

what does this mean? it seems to only happen on one stack...? 🤔

@bitter-oil-46081 did 2744 not fix this as well? Error: Cannot call 'requireOutputSync' if the referenced stack has secret outputs. Use 'requireOutput' instead.

right, manual installation fixes it, brew is still serving beta 2

@bitter-oil-46081 thanks for the welcome improvement 🙂

this is such a gross hack 😶https://github.com/terraform-providers/terraform-provider-postgresql/issues/49

wouldn't even be able to connect to two ssl required dbs