Open jdestombes opened 7 months ago
Issue appears with version 3.85.0 Version 3.84.0 upload the VHD correctly
@jdestombes Could you please provide a complete configuration with the storage account and the blob, with values interpolated, so that we can take a further look?
I am also experiencing this behavior on azurerm_storage_account
when running a large config that interacts with a large number of storageaccounts. Running from ADO pipeline agent hosted on Azure Container Instance (Windows) using provider hashicorp/azurerm v3.100.0
, also experiencing this on 3.99
.
│ Error: building Accounts Data Plane Client: retrieving Storage Account Key: listing Keys for Storage Account (Subscription: "2ccab7fe-ac67-49ce-974e-11a8da78fb5f"
│ Resource Group Name: "xxxxx"
│ Storage Account Name: "xxxxx"): Post "[https://management.azure.com/subscriptions/xxxxx/resourceGroups/xxxxx/providers/Microsoft.Storage/storageAccounts/xxxxx/listKeys?%24expand=kerb&api-version=2023-01-01"](https://management.azure.com/subscriptions/xxxxx/resourceGroups/xxxxx/providers/Microsoft.Storage/storageAccounts/xxxxx/listKeys?%24expand=kerb&api-version=2023-01-01%22): HTTP response was nil; connection may have been reset
│
│ with module.hyper_v_arcservers_rg.module.hyper_v_host_resources["hv27"].azurerm_storage_account.xxxxx,
│ on modules\arc-hyper-v-host-resources\storageaccount.tf line 1, in resource "azurerm_storage_account" "arc_hyper_v_host":
│ 1: resource "azurerm_storage_account" "arc_hyper_v_host" {
From another report I've received, this error might indicates the service is rate limiting your access. I'm not sure if it is intermittent, but you can probably avoid it by not sending too many requests at once.
From another report I've received, this error might indicates the service is rate limiting your access. I'm not sure if it is intermittent, but you can probably avoid it by not sending too many requests at once.
You are right, set parralelism to 1 solved the issue.
FYI setting parallelism to 1 did not solve it for me. running in azure pipeline
I am also getting the same error "HTTP response was nil; connection may have been reset" for different resources- private endpoint, function app, route tables, proximity placement group with the az api version 1.10 and 1.12.
I´m having same problem trying with azurerm ~> 3.100 and setting paralelism to 1:
module.vm["vm-ubuntu-20"].azurerm_storage_blob.vhd_blob[0]: Still creating... [18m30s elapsed] module.vm["vm-ubuntu-20"].azurerm_storage_blob.vhd_blob[0]: Still creating... [18m40s elapsed] module.vm["vm-ubuntu-20"].azurerm_storage_blob.vhd_blob[0]: Still creating... [18m50s elapsed] module.vm["vm-ubuntu-20"].azurerm_storage_blob.vhd_blob[0]: Still creating... [19m0s elapsed]
╷ │ Error: creating Blob "VM-6.vhd" (Account "Account \"predsacvhd\" (IsEdgeZone false / ZoneName \"\" / Subdomain Type \"blob\" / DomainSuffix \"core.windows.net\")" / Container Name "vmimage"): creating storage blob on Azure: while uploading source file "/home/user/Downloads/VM-6.vhd": writing page at offset 3543965696 for file "/home/user/Downloads/VM-6.vhd": executing request: Put "https://vhd.blob.core.windows.net/vmimage/VM-6.vhd?comp=page": HTTP response was nil; connection may have been reset │ │ with module.vm["dsac-scanning-ubuntu-20"].azurerm_storage_blob.vhd_blob[0], │ on ../../../terraform-vm-linux-module/main.tf line 234, in resource "azurerm_storage_blob" "vhd_blob": │ 234: resource "azurerm_storage_blob" "vhd_blob" {
In other cases I saw that probably setting this could be helpful but not works... Ex: timeouts { create = "90m" }
Is there an existing issue for this?
Community Note
Terraform Version
1.7.5
AzureRM Provider Version
3.98.0
Affected Resource(s)/Data Source(s)
azurerm_storage_blob
Terraform Configuration Files
Debug Output/Panic Output
Expected Behaviour
VHD should be uploaded (issue is not present with azruerm 3.0.0)
Actual Behaviour
Connection Reset
Steps to Reproduce
terraform apply
Important Factoids
No response
References
No response