I have a pulumi stack which includes a RDS cluster, which I tried out serverless instanceClass on. but reverted back to t3.medium, and that works fine, the cluster instances are now db.t3.medium. BUT, every time i run

pulumi up --refresh

now it tries to change the following:

serverlessv2ScalingConfiguration: {
        maxCapacity: 16 => 0
        minCapacity: 2 => 0
    }

instanceClass                     : "db.serverless" => "db.t3.medium"

the update succeeds fine, the instancesclass isnt serverless and havent been for a long time, but its annoying that it reports drift every time. How can I fix this?

Hey guys! Can I create a Pulumi Package as a private repository? Meaning publishing it to a private repo inside my organization?

whats the prefered way to get the same versions of pulumi plugins across different machines?

is it possible in pulumi to tell it to only shut down one or two VMs at a time in a group of VMs? for example, if I'm using it to manage the VMs for a kubernetes cluster to tell it to only take down 2 of 6 machines at a time so the whole cluster isn't down?

Have you seen the new provider boilerplate? It's just amazing! https://www.pulumi.com/blog/pulumi-go-boilerplate-v2/

[This question was asked yesterday in #aws with no replies] I’m creating 3 services in AWS Fargate using

@pulumi/awsx@1.0.1

, but for some reason when deploying an update to their container Pulumi decides to deploy the change to each service one after the other rather than in parallel, so a deployment takes 3x as long as it otherwise might. There’s no dependency between the services AFAICT. Is there any way to avoid it doing this, or to figure out why it’s deciding not to parallelise these updates? Thanks!

hello

how do I get the "aws:region" config in Golang? do I load the "config" object and then read "aws:region"?

Any way to deploy more than 1 stack at the same time?. Can be done with custom providers?

Is it possible to use Stack References without having a pulumi account?

Hey guys, we have a pulumi subscription and I’m just wondering the best avenue for getting support. When I run pulumi preview on my stack, it just spins forever and does nothing.

Hi there, I'm trying tp assign azure synapse roles once the firewall rules are implemented but there is a time gap (about 1 minute) between the finished creation of firewall rules and its taking effect. Hence it's throwing an error - Failure responding to request: StatusCode=403 -- Original Error: autorest/azure: Service returned an error. Status=403 Code="ClientIpAddressNotAuthorized. Here custom timeout cannot be used as it's an component resources. How do I tackle this error. Could some one please help me out with this.

Hi there, is it possible to export a value to pulumi stack outputs in .NET. There is a docs page for stack outputs here, but it only has examples in Typescript, Python, Go and YAML. I can't find anything by googling about it. Is it unsupported? Or are there some docs I am missing. Thanks! 🙏

Hi there, I'm having difficulties overriding the CI detection mechanism in Jenkins. I tried adding PULUMI_DISABLE_CI_DETECTION=1 and JENKINS_URL="" but pulumi refresh is still complaining about

error: --yes or --skip-preview must be passed in to proceed when running in non-interactive mode

Does anyone have an idea?

Hi everyone! I am trying to add annotations for adding recordes to the route 53 private hosted zone in my ingress resource, but it's updating in the public hosted zone instead. Not sure if I am using the right annotations. If anyone knows, can you please let me know or point me to the right resource? Thanks!

Hi there, is there have any Pulumi best practices document like Terraform do https://cloud.google.com/docs/terraform/best-practices-for-terraform

Hello! We’re trying to add new members to the organization, but it’s not being possible since apparently, the admin user has not validated his email, however, when going to the account settings page, we can’t find an option to re-send the validation email. Please see the attached screenshots. Can anyone give me some advice?

Hello, Could anybody tried to create aws glue catalog table with Apache Iceberg Integration. I tried to create using the below format, the table got created with wrong format. DDL which I was able to create in Athena console and successfully created:

CREATE TABLE `t_temp_table_iceberg`
(
`col1` string,
`col2` double,
`col3` string,
`col4` BOOLEAN,
`col5` BOOLEAN
)
PARTITIONED BY(
col1,
col2
)
LOCATION '<s3://bucket-name/iceberg_location/>'
TBLPROPERTIES (
  'table_type'='ICEBERG',
  'format'='parquet'
)

Tried to deploy the same in pulumi but the table is created with wrong format similar like Hive table with partitions and not able to access in Athena( Not an issue with permissions verified). Can anyone help me with it.

t_temp_table_iceberg = aws.glue.CatalogTable(
            resource_name="t_temp_table_iceberg",
            name="t_temp_table_iceberg",
            database_name=self.glue_catalog_iceberg_catalog_db.name,
            parameters={
                "table_type": 'ICEBERG',
                "format": "parquet"
            },
            storage_descriptor=aws.glue.CatalogTableStorageDescriptorArgs(
                columns=[
                    aws.glue.CatalogTableStorageDescriptorColumnArgs(
                        name="col1",
                        type="string",
                    ),
                    aws.glue.CatalogTableStorageDescriptorColumnArgs(
                        name="col2",
                        type="double",
                    ),
                    aws.glue.CatalogTableStorageDescriptorColumnArgs(
                        comment="",
                        name="col3",
                        type="String",
                    ),
                    aws.glue.CatalogTableStorageDescriptorColumnArgs(
                        comment="",
                        name="col4",
                        type="Boolean",
                    ),
                    aws.glue.CatalogTableStorageDescriptorColumnArgs(
                        comment="",
                        name="col5",
                        type="Boolean",
                    ),
                    aws.glue.CatalogTableStorageDescriptorColumnArgs(
                        comment="",
                        name="Substituted",
                        type="Boolean",
                    ),                ],
                location=f"s3://{self.curated_bucket_name}/iceberg_location/",
            ),
            partition_keys=[
                aws.glue.CatalogTablePartitionKeyArgs(
                    name="col1",
                    type="",
                    comment="",
                ),
                aws.glue.CatalogTablePartitionKeyArgs(
                    name="col2",
                    type="",
                    comment="",
                )
            ],
        )

Hi everyone, new to Pulumi but loving the product! Have run into an issue where our AWS API Gateway does not deploy when it has an Authorizer. Found an open issue on Pulumi github https://github.com/pulumi/pulumi-aws-apigateway/issues/17 . Can anyone suggest a work-around and/or a sample where this is working successfully? Thanks in advance! 🙏

Hello, does anyone know Pulumi's max string length for project names and stack names?

Messages and files older than 90 days are hidden

I'd like to suggest moving this community to discord. It's more suited to communities too.

Hello all! I am having some issues, probably something stupid, with setting a project with multiple stacks and loading the configuration from the .yaml file from the respective stack. I have the following stacks: • dev-kms • staging-kms And I have these 3 pulumi config files: • Pulumi.dev-kms.yaml • Pulumi.staging-kms.yaml • Pulumi.yaml The config from dev-kms and/or staging-kms are not being loaded when I pulumi.Config(), only the content com Pulumi.yaml appears. code snippet:

"""An AWS Python Pulumi program"""

import pulumi
import pulumi_aws as aws

stack = pulumi.get_stack()

config = pulumi.Config()

Pulumi.yaml:

name: aws
description: kms configuration
runtime:
    name: python
    options:
        virtualenv: env

Pulumi.dev-kms.yaml:

config:
  aws:region: us-east-1

When I run the code only the configuration from Pulumi.yaml goes to config. How can I "fix" this? What I am doing wrong? Appreciate your attention.

Hi there! Does anyone know how to create Image in GalleryImage in Azure Classic or in Azure Native? I've check +- both, I didn't find any link attribute nor in Image and GalleryImage, and didn't find any mapper class or something.

does anyone use the

pulumi logs

command? This seems really cool and useful but every time i go to use it i get

error: failed to get logs:

* no AWS region found

Hi all, currently our team encountered a need to have “multi stage deployment” from pulumi. An example user flow is to run pulumi up first to setup initial infra, then do some development on the initial infra, and then there is a need to deploy the newly developed resources on the initial infra. The current workaround is to have flags, and manually switch them on or off. Is there a better flow (e.g. setup a separate pulumi project just for the last deployment stage)?

Is anyone aware of something like https://github.com/cycloidio/inframap for pulumi, that can scan our code and generate network/architecture diagrams?

Hi Team. Urgent help, please. Suddenly I got this error 😭

error: could not load plugin for random provider 'urn:pulumi:sg-dev::Az-Trans-AKS::pulumi:providers:random::default_4_9_0': Could not automatically download and install resource plugin 'pulumi-resource-random' at version v4.9.0, install the plugin using `pulumi plugin install resource random v4.9.0`.
Underlying error: error downloading plugin random to file: failed to download plugin: random-4.9.0: 403 HTTP error fetching plugin from <https://get.pulumi.com/releases/plugins/pulumi-resource-random-v4.9.0-linux-amd64.tar.gz>

When creating a resource you can provide CustomResourceOptions where you can specify that it DependsOn another resource. But when you use one of the Get/Invoke methods, to get a value from a resource to use for something else you don't have this options in InvokeOptions. What is the way to go to mark an Invoke() to have a dependency on something that potentially hasn't been created yet?

Hello Everyone, Is there a way to read a local shell script and add those lines to the user_data of the ec2 instance. I am looking for a best pratice to do this. If there happens to be one.

Hello experts, I have an auto-scaling issue for AWS ECS Fargate. Thresh hold is not taken from the pulumi scripts.