sleep (seconds) -> sleepUntil (specific date)
Is there a way to wait to execute something until a specific date without needing to calculate sleep in seconds or use cron?
Something similar to what inngest does with sleepUntil (https://www.inngest.com/docs/reference/functions/step-sleep-until)....
how can PowerShell returns an json array?
Is there a way that powershell return an json arry?
I have written a script, that gets aa json array from an restful api. but when i wrote Write-Output $result (the array of json), the return value of the powershell script is on "]" 😦...
somebody already use elasticsearch a destination?
i want to collect some retfull api sources to elasticsearch, but the elasticsearch dest. is not implementent.
Problem with sleep
Hey guys,
We are facing an issue with sleep in the flow. Scenario is we have a step in the flow which needs to sleep for 6 hours and then execute the next step. Problem is after completing the sleep time flow will shut down. See the example in the screenshot it shows the error "Job Cancelled" and http connection was broke by sanket. But we are not touching the flow when it's sleeping.
My observation is it is always crashing after completing the sleep time....
Scripts missing from flow after update
I've just updated for the first time in a while and now when I try to add a workspace script to a flow I can only see two (of 5) folders.
The rest of my folders and their contents are still visible in the workspace home, and the scripts that had been added to the flow before I updated are still there. When I open any of the "missing" python scripts I'm also asked to save or discard changes despite making no alterations....
Open API Key
OpenAPI says it now uses project based keys instead of user key's. When I create a project based key with all permissions, Windmill is saying incorrect API key provided.. am I creating this key in the wrong place in OpenAI?
Staying in sync
I've read through the docs on git sync and wmill cli sync. I have wmill syncing working. When I make changes via the wmill UI, if I want them in my local copy of the workspace I run wmill sync pull. If I make changes to the local workspace, I run wmill sync push, approve the diff and the changes go back to the wmill workspace storage. If I look at the history of a changed file, I see the history of changes. Now I want this workspace to also be synced to a git repo and I have some questions:
If I add git sync to the workspace, do I also need to clone the repo to work locally or do I just use wmill cli for local work and push changes back to the workspace?
For git sync, what if changes happen to the repo outside of wmill, how do I get wmill to pull the latest code from the repo?...
Multiple azure_blob resources
Is it possible to have more than one azure_blob resource per workspace that can work with the S3 File Uploader app component?
The goal is to have the uploaded files in a different container than the one the first resource points to.
I created a second azure_blob resource and configured the workspace S3 storage settings for it to be used as secondary storage. When using the S3 File Uploader in an app it uses the first azure_blob resource by default and I see no way to change it....
clear worker queue
Is there a way to purge the queue? We have hundreds of stuck tasks and cancel all does not do anything
Google Auth Tokens Exposed?
I am working on a project that exposes some API endpoints but would like to secure it on the server-side. Does Windmill Cloud App expose a way to fetch OAuth tokens/create SSO that I could pass to the API endpoint for user verification?
Bun installer appears to fail during script launch.
I just finished setting up a self-host install of WM. I can run a deno typescript test without problems, but if I run a bun test, the mem just keeps building then eventually I see this:
ExecutionErr: error during execution of the script:
process terminated by signal: Some(
9,...
Python imports
I am trying to import Azure's DocumentIntelligenceClient with the statement:
from azure.ai.documentintelligence import DocumentIntelligenceClient
but keep getting the error: ModuleNotFoundError: No module named 'azure'
...Triggering a script using a synchronous webhook from another script within a self-hosted Docker
Hi all!
Windmill and Docker semi-noob here, running into an issue for the last 2 hours I can't seem to figure out. I am trying to call a script from another script within the same docker container. I am able to get it to work async, but when I try to do it async, it seems to trigger the script, but it runs indefinitely and never returns anything. Nor can I see any logs of relevance. I'm using http://windmill_server:8000 as the base of the route, and feel like the issue is related to them both being on the same docker/network.
If anyone has any insight on how to use syncronous webhooks to trigger scripts I'd be immensely appreciative. Thank you!...
404 resource not found
I am getting the following error. I don't see any reference to a_user anywhere in my code or input/output
http://localhost:37761/api/w/spacename/resources/get_value_interpolated/u/username/flowname/a_user: 404, Not found: Resource u/username/flowname/a_user not found...
auto scaling workers
Did anyone figure out how to get the workers to auto scale depending on workers waiting for work or some other arbitrary metric that’s not just cpu and men?
Query params as input args
Is it possible to use query parameters as input args in a script or flow? Current use case is validating a subscription request to MS Graph which sends a validationToken as a query param with no option to pass it as a header. I can envision plenty of other use cases for this as well especially where requests are from sources out of the Windmill operator's control.
Migrating content to another Windmill instance
I have a corp client that is taking forever to make a decision on how they're going to license Windmill. In the meantime, I've been building their integrations using a free cloud instance. I'm curious, is there a documented process for migrating all content from one Windmill instance to another if that becomes a necessity?
will I be forced into prohibitively expensive paid plans later on?
Have been using Pipedream for a few years. Saw your mention of Pipedream in your compare pages, and it looks encouraging. But I'm a small business owner who likes to fiddle with stuff in my spare time to make running my business easier. A couple of years ago I put a lot of effort into setting up workflows in Pipedream, only last year they changed all their plans and arbitrarily made certain (what you call) Resources (they call them Accounts) part of premium plans only. Which means the workflows ("Flows" in windmill) suddenly locked me out of any editing. They continue to work, but as my business changes (like all businesses do) I found I could not edit the workflows anymore to update them. This is pretty annoying - letting me put all that effort in only to try to force me onto a higher paid plan because I have so-called "premium" resources involved. And I know that at their end, the effort required to make a call to Gsheets is exactly the same as making one to Mailchimp or Xero or Airtable or Woocommerce or Stripe or AWS SES. All of which I use. They are just API calls and it leaves a bad taste in my mouth when they try to convince me that I should suddenly pay more for some of them but not others.
To be clear, I am not looking to get "perpetually free" service. Windmill looks really cool and as long as I am aware up front of what the costs will be, I'm happy to start putting in the work. I just don't like to be told one thing then forced into a different direction later on.
I would be happy to pay for things like "increased speed" or for services that can outright replace some of the tools I currently pay for (i.e. redirect the payments from that service to you), but not for arbitrarily restricted functionality....