v5.4.5 Release Notes

Release Dates

  • 5.4.5-1 Apr 13 2022

  • 5.4.5-2 Apr 21 2022

5.4.5-2 Updates Click to Expand/Hide

5.4.5-2 contains the following updates not included in 5.4.5-1:

Agent Stats

  • 5.4.5-2 Fixed 5.4.5-1 Agent stat regression with Memory and Compute statistics when using custom/non-system Instance Type Layouts.

Hosts

  • 5.4.5-2 Convert to managed: Fixed Instance Link not displaying on Host detail page after being converted to managed for some Windows server types.

Kubernetes

  • 5.4.5-2 Fixed EKS and GKE sync issue where Morpheus attempts to delete a Cluster after it has been deleted in the Cloud

  • 5.4.5-2 Fixed IP address assignment when provisioning Morpheus Kubernetes Clusters (MKS) using IP Pools

NSX-T

  • 5.4.5-2 Fixed 5.4.5-1 regression preventing API NSX-T segment creation

Tasks

  • 5.4.5-2 Fixed cypher lookup for Ansible and Python tasks


Important

Database indexes added for account_usage and metadata_tag tables. Customers with very large account_usage and/or metadata_tag tables (10 million+) may experience slower initial morpheus-ui loading times and additional database load after upgrading to 5.4.5 while the indexes are being added.

Note

Items prepended with 5.4.5-2 are added in 5.4.5-2. Items appended with 5.x.x are also included in that version.

New Features

API & CLI

  • Instances endpoint now has Terraform Apply and Terraform Refresh as the Apps endpoint already did

  • Software license functionality (Administration > Settings > Software Licenses) is viewable from Morpheus API and CLI

Clouds

  • Improvements made to Cloud details pages seen for users given read-only access to a particular Cloud

Clusters

  • Add tags when provisioning a cluster or adding a host 5.5.0

  • Morpheus Cluster type Combo Cluster renamed to KVM/Docker Cluster

Credentials

  • Stored Credential sets (Infrastructure > Trust) can now be used with some Option List configurations, just as authenticating against a REST API for REST-sourced lists

  • Stored credentials can now be used with HTTP-type Tasks and for access to remote execution for other Task types

  • Users can now integrate all Cloud types using stored credentials (Infrastructure > Trust) and Morpheus will filter the list of selectable credentials to only types supported by the target cloud

Currency

  • Morpheus now supports Argentine Peso (ARS) currency

Cypher

  • Added Cypher Access Policy type to enable granular control over access to List, Read, Write and Delete on arbitrary Cypher paths for Global, Role, and User scopes

EfficientIP

  • EfficientIP SOLIDserver IPAM plugin is now available on request. Can be added to Morpheus (Administration > Integrations > Plugins) to add the new IPAM integration type (Infrastructure > Network > Integrations) to the appliance

F5

  • Added F5 Policy and Policy Rule creation & management, including http forwarding policies

  • Added Partition sync & selection

Github

  • Updates made to ensure full functionality of Github integrations after Github deprecated SHA-1

Instances

  • Greenfield workloads can no longer be removed from Morpheus without also removing the underlying infrastructure

  • Removed the edit action (pencil icon) from the Network tab of the Instance detail page

  • The Instances list page can now be filtered to show only the Instances owned by the current user. Click the gear icon at the top of the list to edit output fields and filter criteria

KVM

  • The host-local image pool (morpheus-images) now has automatic purging behavior to ensure the allocation doesn’t get too full. At 80% full, images will be purged beginning with the oldest accessed cached volumes until 50% capacity is reached

  • When provisioning to KVM hosts and selecting plans with custom CPU and memory parameters, configured maximum CPU and memory values (if any) on the plan are now honored 5.5.0

Kubernetes

  • CPU, Memory, Storage and Network metrics shown for pods running on a connected Kubernetes cluster

  • Pricing for AKS, EKS, and GKE controllers and workers is now displayed on the review tab of the provisioning wizard

Load Balancers

  • AVI Load Balancer has been renamed NSX Advance Load Balancer to reflect the renaming of the technology itself

NSX

  • Distributed firewalls for NSX-T are now accessible to Subtenants when an NSX-T integration and distributed firewall has been shared from the primary Tenant 5.4.4

OpenStack

  • The version picker for OpenStack Clouds has been removed as it did not need to be set in most cases and created confusion

Policies

  • Added Policy type to limit number of virtual servers on a network

  • Added Policy type to limit the number of members in each load balancer pool within the policy scope

  • Additional scopes added for max load balancer pool policies. Global, Cloud, and User-scoped Policies can now be created

Python

  • Added local workspace path for Python task execution to resolve slow venv execution when /var/opt/morpheus/morpheus-ui is on nfs

Security

  • CVE-2021-30129 Upgrade sshd-core to version 2.7.0

  • Embedded Elasticsearch jackson-databind upgraded to 2.13.2.1. (CVE-2020-36518)

ServiceNow

  • After exposing a Morpheus Catalog Item to ServiceNow, the default workflow can be set and edited without resetting to the default after the nightly sync

  • In most cases, default Input values and help blocks are synced over for exposed Catalog Items as they would be if the Catalog Item were provisioned from Morpheus

  • The “title” value for the Morpheus category of the Morpheus plugin section of ServiceNow can now be updated without reverting back on the next daily sync

Trust Integrations

  • The Venafi integration, which was previously deprecated, has been removed from the product

UI

  • Database optimizations added for usage and tag tables to improve application performance

  • On Instance delete, the “Preserve Backups” option is only shown when backups are enabled for the Instance

Usage

  • “Usage Retainment” setting added to Administration > Settings. Determines how long to keep usage records in database. Retainment period is not set by default. Usage records will remain indefinitely like prior releases if Usage Retainment is not set.

VMware

  • Added “Enable Storage Type Selection” to VMware vCenter Cloud integrations. If selected, storage type (thin, thick-lazy zero, thick-eager) can be selected at provision time

XaaS

Fixes

Agent Stats

  • 5.4.5-2 Fixed 5.4.5-1 Agent stat regression with Memory and Compute statistics when using custom/non-system Instance Type Layouts.

API & CLI

  • api/apps/:id/prepare-apply endpoint has been restored to Morpheus API. It was removed in a previous release

  • A generic warning is now returned when attempting to POST to /provisioning/apps using a Blueprint ID the user cannot access

  • Fixed an issue that caused VMs to go into an unknown state when activating or deactivating clusters via Morpheus API

  • Fixed an issue that caused errors to be thrown when applying Morpheus IP Pools to GCP networks via API

  • Fixed issue with “incomplete configuration” error when updating catalog items via API

  • Fixed Morpheus API-type created via API and shared with a subtenant containing Primary Tenant objects in some cases

  • When simultaneously adding/updating tags and customOptions values on an Instance, existing tags are no longer wiped out. The new tag is simply appended to the existing tags as expected

  • Morpheus API and CLI can now be used to create Cloud-scoped Policies targeted to Clouds which are private to the Subtenant

Amazon

  • Cached and expired STS keys are now cleared properly which prevents authentication issues

Ansible Galaxy

  • Improved cleanup of Ansible Galaxy collection caches to prevent unnecessary storage use

Ansible Tower

  • Fixed an issue that caused the option to disable Ansible Tower at provision time not to work properly

  • Fixed an issue that prevented changes made on the Ansible Tower side not to be synced back to Morpheus

Ansible

  • Fixed an issue that could cause Ansible Tasks in the Provisioning Phase of a Workflow to fail

Apps

  • Resolved issue with ${instance.name} variable not evaluating for VIP Hostname

Bluecat

  • Fixed an issue that prevented removal of Bluecat integrations

Blueprints

  • Editing App Blueprints and provisioning Apps to target Clouds named with their FQDN now works properly

Budgets

  • Fixed an issue that prevented Subtenant users from creating Cloud budgets

  • Fixed incorrect time period labels on multi-year budgets set on quarterly scale with a configured custom fiscal year

Catalog

  • Added executions section to Catalog inventory pages so users can better confirm success of workflows run against multiple Instances and servers

  • Fixed an issue causing a permissions error to be displayed after ordering a Catalog Item if the user did not also have access to see the Inventory which is where the user was redirected after completing a Catalog order

  • Removed “copies” parameter from the JSON config body for Catalog Items as this concept is intended only for provisioning executed via the full provisioning wizard

Clouds

  • Users can no longer advance to the configuration step of the Add Clouds modal without first selecting a Cloud type

Dashboard

  • Fixed issues with log counts and graphs on the Dashboard page (Operations > Dashboards) which caused counts to stay at zero

Distributed Worker

  • Fixed an issue that prevented deletion of distributed workers via Morpheus UI

F5

  • Fixed an issue that caused Instances to become inaccessible when provisioned with F5 load balancer and floating IP address

Google Cloud (GCP)

  • Fixed an issue that caused the console to become inaccessible for GCP Instances using private IP addresses

Hosts

  • 5.4.5-2 Convert to managed: Fixed Instance Link not displaying on Host detail page after being converted to managed for some Windows server types.

Identity Sources

  • Fix db lock issue causing user creation failure when using Custom External SSO Identity source

Infoblox

  • Infoblox host records are no longer created with the zone name in the “name” field

Inputs

  • “No Options Found” message on empty Select List Inputs is no longer selectable to prevent confusion and satisfying required Inputs

  • Fixed a display issue when editing Instances that caused updated Input values to revert and the Edit Instance modal not to close after saving changes despite the changes being saved in the background

  • Improved handling of dependent Inputs which are shown on the Edit Instance modal when making changes on the Instance detail page

Instances

  • Fixed an issue that caused the “Cloud” link (to the target Cloud detail page) to appear on Instance detail pages for users with “None” access to “Infrastructure: Clouds” under certain conditions

KVM

  • Fixed an issue that could cause failures when reconfiguring KVM Instances to add or remove secondary network interfaces

Kubernetes

  • Fixed an issue that could cause failures when adding container nodes to a Kubernetes cluster

  • Fixed an issue that could cause failures when provisioning MKS clusters using Plans with a custom storage range

  • 5.4.5-2 Fixed EKS and GKE sync issue where Morpheus attempts to delete a Cluster after it has been deleted in the Cloud

  • 5.4.5-2 Fixed IP address assignmnet when provisioning Morpheus Kubernetes Clusters (MKS) using IP Pools

  • Health checks for container services provisioned to Kubernetes Clusters now work as intended

Layouts

  • “Enable Scaling (Horizontal)” option now works on individual Layouts even when the Instance Type is configured to disable it

Monitoring

  • The breadcrumb link from a Monitoring Group detail page back to the Monitoring Groups list page now works correctly

NSX-T

  • 5.4.5-2 Fixed 5.4.5-1 regression preventing API NSX-T segemnet creation

  • Fixed an issue causing member groups on NSX-T load balancer pools not to be saved and persisted properly

  • Fixed an issue that could prevent editing and resaving NSX-T segments with errors

  • Fixed an issue that prevented selection of NSX-T load balancer SSL profiles in App Blueprint and App wizards

  • Subtenant users with sufficient Role permissions can now drill into NSX-T routers shared from the Primary Tenant

Network

  • Corrected an issue that could result in Instances having multiple primary NICs

  • Fixed an issue that prevented removal of IP Pools from a subnet

Node Types

  • System-default scripts are no longer selectable on user-defined Node Types since they lack the inputs needed to work properly and weren’t intended for use outside of the default Node Types

OpenStack

  • Fixed an issue causing errors when resizing network for OpenStack Instances via reconfigure

  • Fixed an issue related to OpenStack floating IP Pools not respecting associated network permissions

Option Lists

  • Fixed an issue that caused the Option List size value (on the list page for Option Lists) not to be reported corrected in some cases

Plans & Pricing

  • “Show Pricing” setting (Administration > Settings > Provisioning) is now honored in Subtenants as well

  • Snapshot price sets can now include ‘datastore’ price types in addition to the required storage price type (‘Disk Only’)

  • Software prices are now included in computed prices

Provisioning

  • Fixed an issue that caused the provisioning wizard not to work properly when only one Instance Type and Layout was exposed to a Subtenant user

Roles

  • The Tools menu will now be shown for users whose Role only gives access to VDI Pools and nothing else under Tools

  • Users with no permissions to “Library: Virtual Images” can now see and add additional disks on cloned Instances

Security

  • Fixed permission issue with /library/services api endpoint

  • Fixed permission issue with /settings/software-licenses api endpoint

  • Layout descriptions are now limited to 1,000 characters for security and performance reasons

  • Password reset email links are now active for 30 minutes for security reasons. Previously they were active for seven days

  • Security enhancements added to close potential XSS and CSRF attack vectors

ServiceNow

  • Fixed an issue with custom ServiceNow CMDB class mapping

  • Removed the “enabled” flag displayed for exposed Catalog Items on the ServiceNow integration detail page as there is currently no concept of enabling or disabling exposed Catalog Items

Storage

  • Fixed an issue provisioning uploaded images with many disks

Tasks

  • 5.4.5-2 Fixed cypher lookup for Ansible and Python tasks

Terraform

  • Added capability to add tfvar secret to Terraform Layouts using Morpheus API and CLI

UI

  • Fixed a display issue that caused App Blueprint configuration windows to be compressed when the App Blueprint was given a very long name

Usage

  • Fixed usage issues associated with Snapshot, Virtual Image and Load Balancer price types. Datastore ID property added to disk price types API queries

VDI Pools

  • Fixed broken custom logos for VDI apps

VMware

  • Fixed Resource Pool folders not syncing in order which caused unexpected behaviors

  • Fixed an issue that could lead to duplicate SCSI controller and volume external IDs which created additional problems

  • Fixed issue that could cause disk layout to be mismatched in Morpheus compared with the vCenter console

Wiki

  • Improved sync of Wiki information between the main Wiki section (Operations > Wiki) and the Wiki tab of Instance detail pages

Workflows

  • Added apiAccessToken for “configuration” workflow phase

vCloud Director

  • Fixed an issue that could cause provisioning failures to vCD networks created in Morpheus

  • VDCs associated with a private vCD Cloud are no longer visible in Subtenants

Appliance & Agent Updates

Appliance

  • ui['jobs_enabled'] = true/false config setting added to morpheus. This option disables the appliance jobs service on the appliance node when set to false. This should be disabled only when configuring jobs to run on specific app nodes in HA environments.

Deprecations

A10

  • The A10 Load Balancer type has been disabled, and will no longer be an option when adding new Load Balancers. Contact Morpheus if you need to re-enable A10 Load Balancer option. This does not affect existing Load Balancers.

AppDynamics

  • The AppDynamics integration, which was previously deprecated, has been removed from the product

Clouds

  • Improvements made to Cloud details pages seen for users given read-only access to a particular Cloud

  • Dell, HPE (NOT HPE Oneview), Supermicro and Cloud Foundry Cloud Types are now disabled by default. They can still be re-enable in Appliance Settings. Does not affect existing Clouds.