Windmill

W

Windmill

Join the community to ask questions about Windmill and get answers from other members.

Join

help

general

changelog

I'm trying to restore a self-hosted instance but it's seen as a new instance

I have a backup made by docker-volume-backup which just archives the PostgreSQL bind-mounted volume. I also have a self-made backup using pg_dumpall. When I restore with either method, the server seems to think that this is a brand new instance. How do I get my old state back?

Error while fetching flow status

Hello, I recently encountered an error I can't quite make heads nor tails on. The following error below occurs after the myorgOrderC step around 5pm UTC in the attached flow. ```json {...

msgraph python cli

Hello. Does someone alredy succeeed to used msgraph python cli in a windmill script ? Got the error "Starting with v5.0.0, the 'azure' meta-package is deprecated and cannot be installed anymore. Please install the service specific packages prefixed by azure needed for your application."...

Github app button missing in some cases when creating github repository

Hi, I have multiple prod/non prod instances running the same version of windmill EE. For some reason, on one of them the github app button is not showing when I attempt to setup a git repository for Git sync. Wondering where I should look to figure out why. On the instance that doesn't show the button, it shows it when I attempt to create the resource in the "admins" workspace, but not my "non-prod" workspace....
No description

Best practices and guidelines

Hi Team, I am new to windmill.dev, but I have loads of experience as a data engineer using other orchestration tools like Airflow. In my previous experience, we would always keep the heavy data loads and data processing outside of Airflow like in aws batch, spark and orchestrate the jobs via Airflow as a workflow engine. Is it the same practice with Windmill.dev is there any docs or guidelines that explains best practices and what to do and what not do at enterprise scale ? Eg: processing millions of records per hour...

Error after Big For-Loop

Hi everybody! I have a flow with a For-Loop that works great if the loop gets 50-100 items. But if I let it run with 2000 iterations that takes more than a half an hour to complete, I get the following error when exiting the loop : ...

text input

i have a quick question on how to stop caching of a Text Input. When i switch to the next request, the content from the previous Text Input is carried over into the new request. This happens even though each request is supposed to have a fresh, empty Text Input....

Base64 Decoding

I've got a simple script that retrieves file/image contents from S3, and returns the base64 decoded contents as just a raw application/octet-stream HTTP response. When this is run locally (decoding a 588 kB string) the output is a 441 kB valid PNG, but when Windmill runs the same code, it returns 422 B of corrupted & invalid text. Is there potentially an issue post-processing/handling binary/application octet-stream HTTP responses from Windmill flows?
No description

Salesforce Queries Returning Invalid Info When Run on Windmill

Hello! I've got a very small script that uses the simple_salesforce Python library to query my company's accounts. When I run that code locally the name on an account I care about is (for the purposes of privacy) "Infinity Enterprises", but when running on Windmill it returns " null Enterprises"; the word "Infinity" gets replaced with " null " inside that string field. Do y'all have any idea what could be causing this?

[Typescript BUN]: Listing all objects within S3 Bucket

I am trying to get a list of all ".pngs" within a s3 bucket folder which then feeds the dropdown selector we have on Windmill. Right now it's hardcoded where we have a background runnable and a mapping like this. I have already connected the resource and am able to get region, bucketname, etc using
await wmill.getResource('u/path/brisk_s3_aws_oidc')
await wmill.getResource('u/path/brisk_s3_aws_oidc')
...

[Typescript BUN]: Getting SSL certificate error even though I have added the CA as volume mount.

Logged into the container and saw that the certificate is getting added to the truststore and when running curl commands it is working. But when executing the script from Windmill it is failing. Below is the docker-compose.yml windmill_worker:...

My code is gone after going Settings and picking another script kind

Hello. I'm new to wmill. I have tried some code and wanted to explore the UI. I've clicked Settings in top right, then picked "Trigger" as "Script kid" just for exploring. However when I've clicked Action as script kind again, my original code was gone. I wonder if I made a mistake or this is some UX issue?

Flow being canceled by monitor

Why my flow keep being canceled
No description

Setting default_job_timeout via instance_settings.yaml

Hello, I am unable to set the instance-level job timeout via my instance_settings.yaml . I created a github issue for this over the weekend (https://github.com/windmill-labs/windmill/issues/6281) . Pleae let me know if there is something wrong with my approach or if this is indeed a Windmill bug. Thank you!...

Saving to App State from File Input.

I am trying to create a CSV previewer in the windmill app editor (more steps later that is just step 1)... and I'm thinking my process should be that when a file is uploaded to the "file input" component, a background runnable will update state.csvContent to the decoded csv string, and then my preview text box watches state.csvContent and updates appropriately. The issue i'm encountering is that I can't seem to get state to update. In my background runnable, I have passed state as an input. ```js export async function main(csv_file: string, state: any) {...

mssql returns double-nested results

MSSQL: Returns [[{...}, {...}]] (array wrapped in another array) PostgreSQL: Returns [{...}, {...}] (direct array of objects) is this expected bahaviour please? (using agent windows worker)...

For Loops hangs if any iterations fail

Hi! Thank you for the great tool! I was testing the following flow: - it fetches rss articles - processes articles in a for loop - fetches article content...

WINDMILL TOO BIG

I'm trying to pass 600kb of bytes (tried both parquet and pickle) between steps in a flow. But interface says WINDMILL TOO BIG and when I try to load that as a variable into the next step it seems to me like it tries to load "WINDMILL TOO BIG" as a string. I though there was a 50MB cap on this. Could someone help me?

We are setting up workspaces to share with our teams is it possible to use OKTA AD-Group to give acc

We are setting up a self-hosted instance and we are integrating okta. Is it possible to use AD Group to give access to workspaces ?...

Including query params

Do HTTP routes support this? Trying with the following: ?key=1234&include_query=key But it's not being used and shows the key input to be empty....