witty-diamond-30724
07/19/2021, 6:50 PMbetter-shampoo-48884
07/19/2021, 7:44 PMλ pulumi stack import --file c:\tempstack\stage.k8s.deployment.json
error: could not deserialize deployment: unexpected end of JSON input
Any other hints on what to do? It's a 25mb stack..bumpy-businessperson-79451
07/19/2021, 8:12 PMechoing-twilight-40421
07/19/2021, 8:17 PMable-pager-97491
07/19/2021, 8:25 PMlimited-rain-96205
07/20/2021, 12:33 AMcreamy-raincoat-13266
07/20/2021, 12:54 AMcreamy-raincoat-13266
07/20/2021, 12:55 AMerror: getting secrets manager: secrets (code=Unknown): AccessDeniedException: The ciphertext refers to a customer master key that does not exist, does not exist in this region, or you are not allowed to access.
status code: 400
Look to be related to something like this https://github.com/pulumi/pulumi/issues/4665 - has anyone been able to resolve this?fierce-market-67222
07/20/2021, 8:57 AMHello,
i got a problem with the pulumi_gcp, the existing DNS records are not overwritten anymore…pulumi fails with resource is existing…
It was working on pulumi 2 and its not working anymore in 3 … i need to make sure to delete all the existing records sets before…
Does anybody has the same issue?
> **Note:** The provider treats this resource as an authoritative record set. This means existing records (including
the default records) for the given type will be overwritten when you create this resource in the provider
sticky-match-71841
07/20/2021, 11:14 AMrough-dinner-7791
07/20/2021, 2:35 PMgcp.container.Registry
then docker.image.Image
to build it. Every time i make a pulumi up, it's took a looooong time to compile my dockers and then to repush then on my registry, even if their is no change.
Pulumi don't use cache ? If i do it manually using docker build
command, it's instant.
Thank'sgorgeous-minister-41131
07/20/2021, 8:32 PMprehistoric-london-9917
07/20/2021, 8:45 PMinterface PeopleArgs {
name: string
age?: number
}
And a config file that looks something like:
config:
things:people:
- name: Peter Parker
age: 40
- name: Mary Jane
- age: 38
And in my Pulumi program I load the config:
let config = new pulumi.Config("things");
let people = config.requireObject<PeoplegArgs[]>("people");
The first two should be fine. But I’d expect the third element to raise an error because it doesn’t satisfy the PeopleArgs
interface (it’s missing name
). Instead, it seems that Pulumi puts undefined
there.
Is this expected behaviour or a bug? And has anyone who’s encountered this found a workaround?enough-truck-34175
07/20/2021, 8:57 PMcareful-pager-92424
07/21/2021, 10:09 AMaws:cognito:UserPoolDomain (dev.myapp.user-pool-client.web-app.domain-name):
error: deleting urn:pulumi:dev::myapp::aws:cognito/userPoolDomain:UserPoolDomain::dev.myapp.user-pool-client.web-app.domain-name: 1 error occurred:
* Error deleting User Pool Domain: InvalidParameterException: The domain is not configured for this user pool.
Google search came up with this but being new to IaaC world I really don't understand difference between pulumi state delete
and pulumi state unprotect
.
What can potentially go wrong when I run any of these commands? I really wouldn't like to lose my user pool with all users inside.
Thanksminiature-leather-70472
07/21/2021, 2:50 PMworried-hydrogen-32482
07/21/2021, 7:15 PMnumerous-thailand-80976
07/21/2021, 7:37 PMadamant-fountain-88722
07/21/2021, 8:07 PMquiet-hairdresser-18834
07/21/2021, 9:51 PMbored-monitor-99026
07/22/2021, 3:40 AMvenv
and updated pulumi.yaml
, using`pipenv` instead. with everything installed, now when i do pipenv run pulumi up
, i got
error: Exception calling application: Program run without the Pulumi engine available; re-run using the `pulumi` CLI
how to fix this error? thanks!kind-park-65452
07/22/2021, 6:55 AMs3_obj = aws.s3.get_bucket_object(bucket=self.bucket_name, key=self.csv_object_key)
This call is getting fail with an exception I couldn't find the root cause:
venv/lib/python3.9/site-packages/grpc/_server.py:457: RuntimeWarning: coroutine 'invoke.<locals>.do_rpc' was never awaited
return None, False
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
pulumi-python:dynamic:Resource (app-map-migrated-tag):
error: Exception calling application: There is no current event loop in thread 'ThreadPoolExecutor-0_0'.
At first the pulumi_aws package version was 3.33.0 and I even tried to upgrade it to the latest (4.13.0) but no luck.
Any idea here what could be the issue?freezing-quill-32178
07/22/2021, 9:28 AMHelm
and pulumi.StackReference
, particularly slow pulumi up
… for example stack that usually took ~2 mins now take 5+ 10, some Helm stacks never finish so had to be killed as well (which lost the actual stack.json state and had to be recreated from backup and cleaned up)
The Helm resources are marked as deleted (and actually deleted on the k8s cluster) but the pulumi up
is stuck and never finishes), inspecting the stack json the resources are still present in the pending_operations
.
I’m using cert-manager
helm chart for example, failed during install and during uninstall as well
Also seeing a lot of
I0722 11:14:27.696360 74884 request.go:655] Throttling request took 1.07776852s, request: GET:https://[REDACTED]/apis/flowcontrol.apiserver.k8s.io/v1beta1?timeout=32s
I0722 11:14:38.496586 74884 request.go:655] Throttling request took 1.098435921s, request: GET:https://[REDACTED]/apis/admissionregistration.k8s.io/v1?timeout=32s
great-sunset-355
07/22/2021, 9:41 AMname_prefix
- I found several existing issues about this so I'm not sure I should even use it.
One of the issues: https://github.com/pulumi/pulumi/issues/6155
With pulumi resource you have to provide resource_name
and if nothing else is set, this is used in the name of resource.
eg. SNS Topic:
sns.Topic(
"my-topic",
)
Results in the resource name my-topic-1234
. - this at least gives the reference between actual resource and pulumi resource
However when I add the name_prefix
it overrides the resource name completely.
sns.Topic(
"my-topic",
name_prefix='my-prefix-'
)
Results in the resource name: my-prefix-20210722092833209300000002
Is there any documentation about this behaviour? I assume this is inherited from terraform provider and may be resoveld with pulumi native provider which I heard about being in development, correct?
Is it better for now to use pulumi resource name
instead of name_prefix?great-sunset-355
07/22/2021, 11:08 AMregister_outputs
vs pulumi.export()
that provides stack outputs.
I have 2 projects
core
and app
inside core
there is a ComponentResource
class MyTopic(ComponentResource):
"""My SNS Topic"""
def __init__(self, config: TopicModel, opts: ResourceOptions = None):
super().__init__(
"stekz:components:aws:sns-topic",
config.name,
None,
opts,
)
self.topic = sns.Topic(
"component_topic",
name_prefix=config.name_prefix,
)
self.register_outputs(
{
'component_topic_name': self.topic.name,
}
)
# now we instantiate the resource
MyTopic(
config=TopicModel(
name='core-imported',
name_prefix='cp'
)
)
By doing so I expected to be able to refer to MyTopic.outputs.component_topic_name
from app
however that does not seem to be possible because all I can do is to use the StackReference
. But stack reference can only refer to outputs from pulumi.export('key', 'value)
So instead of using register_outputs
I'd have to do this:
t= MyTopic(
config=TopicModel(
name='core-imported',
name_prefix='cp'
)
)
pulumi.export('component_topic_name', t.name)
Which I consider quite ineffective in case this has to be done on a larger amount of resources. Of course, I can wrap it in the function and invent my own convention for this but my point is that I should not have to.
I see this as a problem inherited from terraform which suffers a very similar problem. https://github.com/hashicorp/terraform/issues/8554
I wonder if these issues is somehow related to my problem: https://github.com/pulumi/pulumi/issues/2653, https://github.com/pulumi/pulumi/issues/5941
thank you.
I'm curious if there is any specific limitation that does not allow referencing to the outputs of another stack the way I'd expect or is just the result of the initial implementation where pulumi team took ideas from existing solutions?
Thank youlittle-market-63455
07/22/2021, 12:41 PMaws:someProperty: 'a'
aws:somethingElse: 'b'
vs.
aws:
someProperty: 'a'
somethingElse: 'b'
I feel this helps clear the confusion, at least mine, around reading nested configs and allows you to fold/unfold related options
Something also related to this, why isn't the configuration key types inferred from the config file rather than specified explicitly. As in why requireString()
and requireObject()
and so on instead of just require()
and get
. I feel the config file already all has the type and secret semantics that you need for that?mysterious-australia-14256
07/22/2021, 3:35 PMsteep-toddler-94095
07/22/2021, 9:27 PMrefined-kitchen-10197
07/23/2021, 3:00 AMwet-sunset-4939
07/23/2021, 4:27 AMwet-sunset-4939
07/23/2021, 4:27 AM