I am using Designer in Azure ML studio to build a prediction model. I keep getting the following error: "Deploy: Service status is healthy, but test call return 'BadGateway'" when I deploy a real-time inference of the model, even after using the designer model structure from the Diabetes tutorial. The model publishes and deploys "successfully" (endpoint and keys are created) but it can't be consumed.
How can I resolve this issue?
Related
All, when running a build pipeline using Azure Devops with ARM template, the process is consistently failing when trying to deploy a dataset or a reference to a dataset with this error:
ARM Template deployment: Resource Group scope (AzureResourceManagerTemplateDeployment)
BadRequest: The document creation or update failed because of invalid reference 'dataset_1'.
I've tried renaming the dataset and also recreating it to see if that would help.
I then deleted the dataset_1.json file from the repo and still get the same message so it's some reference to this dataset and not the dataset itself I think. I've looked through all the other files for references to this but they all look fine.
Any ideas on how to troubleshoot this?
thanks
try this
Looks like you have created 'myTestLinkedService' linked service, tested connection but haven't published it yet and trying to reference that linked service in the new dataset that you are trying to create using Powershell.
In order to reference any data factory entity from Powershell, please make sure those entities are published first. Please try publishing the linked service first from the portal and then try to run your Powershell script to create the new dataset/actvitiy.
I think I found the issue. When I went into the detailed logs I found that in addition to this error there was an error message about an invalid SQL connection string, so I though it may be related since the dataset in question uses Azure SQL database linked service.
I adjusted the connection string and this seems to have solved the issue.
I have an Azure Data Factory with a pipeline that I'm using to pick up data from an on-premise database and copy to CosmosDB in the cloud. I'm using a data flow step at the end to delete documents that don't exist in the source from the sink.
I have 3 integration runtimes set up:
AutoResolveIntegrationRuntime (default set up by Azure)
Self hosted integration runtime (I set this up to connect to the on-premise database so it's used by the source dataset)
Data flow integration runtime (I set this up to be used by the data flow step with a TTL setting)
The issue I'm seeing is when I trigger the pipeline the AutoResolveIntegrationRuntime is the one being used so I'm not getting the optimisation that I need from the Data flow integration runtime with the TTL.
Any thoughts on what might be going wrong here?
Per my experience, only the AutoResolveIntegrationRuntime (default set up by Azure) supports the optimization:
When we choose the data flow run on non-default integration, there isn't the optimization:
And once the integration runtime created, we also couldn't change the settings:
Data Factory documents didn't talk more about this. When I run the pipeline, I found that the dataflowruntime won't work:
That means that no matter which integration runtime you used to connect to the dataset, data low will always use the Azure Default integration runtime.
SHIR doesnt support dataflow execution.
I have created VMSS using custom image. I have hosted web application build in .Net MVC on VMSS. I have configured CI/CD from Azure DevOps by referring following https://learn.microsoft.com/en-us/azure/devops/pipelines/apps/cd/azure/deploy-azure-scaleset?view=azure-devops .
It is showing error D:\a\_temp\1575277721063\packer\packer.exe failed with return code: 1 . Any suggestion/recommendation is appreciated.
Below is some failed commands in Log:
1. azure-arm: resources.DeploymentsClient#CreateOrUpdate: Failure sending request: StatusCode=200 -- Original Error: Long running operation terminated with status 'Failed': Code="DeploymentFailed" Message="At least one resource deployment operation failed. Please list deployment operations for details. Please see https://aka.ms/DeployOperations for usage details."
2. Some builds didn't complete successfully and had errors:
2019-12-02T09:57:31.5222618Z --> azure-arm: resources.DeploymentsClient#CreateOrUpdate: Failure sending request: StatusCode=200 -- Original Error: Long running operation terminated with status 'Failed': Code="DeploymentFailed" Message="At least one resource deployment operation failed. Please list deployment operations for details. Please see https://aka.ms/DeployOperations for usage details."
3. 2019-12-02T09:57:31.5222868Z ==> Builds finished but no artifacts were created.
Since the error message (DeploymentFailed) from the pipeline is a generic one, it would be tough to investigate the issue without looking at the underlying logs or your pipeline details.
For troubleshooting it further, please try the following:
View deployment history with Azure Resource Manager as mentioned in the error message itself.
Gather logs to diagnose problems, such as debug/verbose pipeline logs, worker/agent diagnostics logs etc..
Look at some common issues and resolutions if it helps.
Send feedback and report problems through the Developer Community for Azure DevOps.
We have a setup of WSO2 API Management in a distributed pattern (pattern-3) in Kubernetes. We are using a PostgreSQL DB which is running outside the Kubernetes cluster for all the databases.
I have published some APIs in the publisher and am able to invoke them from the store.
I had to make a change in api-manager.xml for the API Publisher and API Store configmap files and recreated the pod. When the pods were available, I observed that the APIs that I had published and working earlier are not visible anymore.
I tried to add the same APIs again and it is complaining that the APIs by that name already exists.
Following is the log from the plubisher pod:
[2019-05-16 08:19:38,266] ERROR - APIProviderHostObject Error occurred while adding the document. PizzaShack API Documentation already exists for API PizzaShackAPI-1.0.0
[2019-05-16 08:19:38,273] ERROR - docs:jag org.wso2.carbon.apimgt.api.APIManagementException: Error occurred while adding the document. PizzaShack API Documentation already exists for API PizzaShackAPI-1.0.0
While creating the API again on the Publisher, following error is displayed: "Duplicate API Name"
It clearly seems to be some synchronization issue. How can this issue be fixed?
I had shared the instance of Carbon DB across the components. This was causingthe issue. Using separate instance for each component in disbuted mode solved it
I trained a model using watson machine learning service. The training process has completed so I ran these command lines to deploy it:
bx ml store training-runs model-XXXXXXX
I get the output with the model ID
Starting to store the training-run 'model-XXXXXX' ...
OK
Model store successful. Model-ID is '93sdsdsf05-3ea4-4d9e-a751-5bcfbsdsd3391'.
Then I use the following to deploy it :
bx ml deploy 93sdsdsf05-3ea4-4d9e-a751-5bcfbsdsd3391 "my-test-model"
The problem is that I'm getting an endless message saying:
Checking if content upload is complete ...
Checking if content upload is complete ...
Checking if content upload is complete ...
Checking if content upload is complete ...
Checking if content upload is complete ...
When I check in COS result bucket the model size is ~25MB so it shouldn't be that long to deploy. Am I missing something here ?
Deploying the same model using Python Client API:
from watson_machine_learning_client import WatsonMachineLearningAPIClient
client = WatsonMachineLearningAPIClient(wml_credentials)
deployment_details = client.deployments.create( model_id, "model_name")
This showed me very quickly that there is an error with the deployment. The strange thing is that the error doesn't pop up when deploying with command line interface (CLI).