Compliant cloud data center upgrades

The data center for Compliant Cloud tenants was updated which had some minor impact on performance and connectivity.

Task importing failures

Nuget.org feed update caused importing new Tasks to fail due to Download counts going over Int32.MaxValue when using the v2 feed

This was mitigated by doing manual package installs for dependencies and fixed by nuget.org capping the download count to Int32.MaxValue. Migration to newer NuGet client libraries expedited.

High Agent memory usage on some Agents

Faulty Windows update to virus protection caused Agent memory usage to spike, which may have caused the Frends Agent service to shutdown. Microsoft has since released a fix for it.

FRENDS NuGet.org gallery outage 11.2.2022

Importing Tasks in the UI may fail if it has dependencies which have not yet been imported.

Affected functionality:

• This only affects importing new Tasks or new versions which have updated dependencies, Agents or existing Processes are not affected.

FRENDS Task feed outage 16.11.2020

Myget NuGet service feeds were not accessible.

Affected functionality:

• Using the "Search for packages" functionality on the Tasks page not possible.

FRENDS Task feed outage 30.10.2020

Myget NuGet service feeds were not accessible.

Affected functionality:

• Using the "Search for packages" functionality on the Tasks page not possible.

FRENDS DNS outage 21.10.2020 15.30 - 16.15

The domain registration for frendsapp.com expired. Because of this the management web sites' and Agents' DNS addresses might not have been able to be resolved for while depending on DNS cache expirations.

Affected FRENDS platform functionality:

• Web UI DNS address might not resolve
• Agent DNS address might not resolve (affects HTTP and API Triggers)

Steps have been taken to prevent this from happening again.

FRENDS partial outage 25.06.2019 12.30 - 17.30

One of the Azure Service Bus scale units had its SQL container run at a very high CPU, which caused some of the Service Bus namespaces used by FRENDS to timeout most of the message send attempts. The SQL container had to be manually scaled by Microsoft to mitigate the issue.

Affected FRENDS platform functionality:

• Process execution logging and Agent status reporting delayed
• Process, API and Environment Variable deployments and other configuration updates to Agents not functioning
• Remote Subprocess executions failing

Microsoft is rolling out a change that will ensure SQL autoscaling when encountering higher workloads to prevent a repeat of this issue.

In addition the FRENDS development team is researching an alternative to Service Bus for remote Subprocesses which would most likely be rolled out in the 5.3 release as an opt-in feature. Also we are also developing our communication practices if similar problems arise in the future.