Instances Overview | Pricefx Knowledge Base

Up to version 1.72.0, instances were called clusters.

The Instances view provides a complete list of all instances and their details. This includes account information, technical details, database information, and instance availability data.

The list is fully synchronized with Salesforce (based on the serial number field). If some information is missing in Salesforce or PlatformManager and it is known, it is added automatically in the other system.

When an instance stops reporting, it is deleted – after the deletion is approved in a workflow. This workflow is automatically initiated when the last recorded message from the instance is older than 2 weeks.

You can also export the table as a CSV file.

Prerequisites

Permissions

This section requires the following permissions.

Instance List – read – Required to view the Instances table.
Instance Admin – Required to manage and edit instances.
Partition Activity Log – read – Required to view logs and monitoring dashboards.
Instance Availability Management – Required to hibernate and wake up instances.

Instances Overview Columns

Column	Description
Account Status	Lifecycle stage (e.g., Onboarding, Business Live).
Version	Current software version of the instance.
Provider	Cloud provider (e.g., AWS, GCP).
Check-in Date	Last recorded check-in of the instance.
Instance Overrides	Displays JSON-based configuration overrides applied to the instance (e.g., query timeout, thread settings, cluster flavor).
Location	Deployment region or data center.
Account ID	Customer account identifier.
Name	Instance name.
Account Name	Customer account name linked to the instance.
Type	Environment type (e.g., Production, Non-Production, Cloud).
URL	Direct access link to the instance.
Status	Current runtime state (Running, Stopped, Hibernated, Not Responding). When someone trigger an action (Hibernate, Wake Up, Set Stop Schedule, Remove Schedule), PlatformManager displays in progress icon in the Status column. The following actions are temporarily unavailable until the operation completes: Hibernate Wake Up Set Stop Schedule Remove Schedule This prevents users from starting multiple conflicting operations. When the operation finishes: The Status returns to its normal state (Running, Hibernated, etc.). All availability actions become available again. Note: Deleted and non-responding assets (partitions, instances) disappear from PlatformManager in 24 hours.
Shutdown Date	Scheduled shutdown time (if configured).
Wake-up Date	Scheduled wake-up time (if configured).
Last Login	Latest login across all partitions of the instance.
Release Name	Release label (marketing version).
Last Upgrade	Date and version of the last upgrade.
Next Upgrade	Date and version of the scheduled upgrade.
Major Version Upgrade Duration	Time taken for the last major upgrade.
Minor Version Upgrade Duration	Time taken for the last minor upgrade.
Database Vendor	Database type (e.g., PostgreSQL, MariaDB).
Database Version	Version and distribution (e.g., PostgreSQL 17.5, MariaDB 10.11).

You can also download this table in the CSV format.

Available Actions

Action	Description	Required Permission
Edit Custom Upgrade Times	Configure maintenance windows for major and minor version upgrades.	Instance Admin
Instance Admin	Opens a table where you can manage and view instances. The table columns are: Server Roles (to manage what every node should do; roles can be added/removed) ListProcessor – A node that can (and will) process background jobs such as price list and price grid calculations etc. For distributed calculations this node will hold the master. CalculationSlave – A node that accepts calculation requests as distributed slave. CalculationFlowProcessor – A node that processes CF jobs. CalculationFlowFailoverProcessor – A node that that will assume CalcFlowProcessor if primary fails (totally useless and should not be used in next gen!). PADataLoad – A node that runs PA data load jobs. These nodes require fast disks. EventProcessor – A node that processes (=sends) generates events out. EmailProcessor – A node that sends emails. Any node can create email tasks, but the actual sending to the MTA happens here. Janitor – A node that performs general house keeping. CalculationScheduler – A node that processes the schedule (mainly PA for now, but this scheduling system will be favored over CF in the future). Restrict calculations to partition Restrict PA data loads to partition View information on: Jetty Host, Jetty Port, Last Join Date, Last Seen Date A note on Server Roles: Setting/changing them really only makes sense in the old bare metal environments. In next-gen infrastructure, these server roles still exist, but they work in tandem with the k8s pods (i.e. the infrastructure) and cannot be viewed individually. As such, the below role explanation still applies, but the settings should not be changed really.	Instance Admin
Show Logs	Shows the log. This is available only for instances with Loki (log aggregation system).	Partition Activity Log – read Note: This option is always accessible to Pricefx internal users.
Show Monitoring Dashboard	Opens external link to the Grafana Monitoring Dashboard for the selected partition. This allows users to quickly access monitoring data such as instance health, cluster performance, and related metrics. Not all clusters have Grafana dashboards available. If a dashboard is not available for the selected partition, no monitoring data will be displayed.	Partition Activity Log – read Note: This option is always accessible to Pricefx internal users.
Version History	Shows a table with used IM versions and dates of upgrade. Versions are stored since May 2024, no historical data is provided.	Instance List – read
Hibernate	Hibernates the selected instance(s). Useful for pausing QA/DEV environments to reduce cloud costs. See Instances Overview. Stop/start connected integration instances: If enabled (default) PlatformManager automatically stops connected IntegrationManager instances when the corresponding Core cluster goes offline. This prevents the integrations from sending data to unavailable partitions. See Integrations Table Options \| Automated Stop and Start.	Instance Availability Management
Wake Up	Resume an instance from hibernation. (Displayed only when an instance is hibernated) See also Instances Overview	Instance Availability Management
Set Stop Schedule	Opens the Instances Stop Schedule dialog, where you can define when selected instance(s) should automatically stop and wake up. Two modes are supported: Simple schedule: Define shutdown and wake-up times directly. Custom cron: Use a cron expression for advanced recurring schedules. Additional options include setting the repeat frequency (e.g., daily) and selecting the timezone. The instance will automatically restart after the defined stop period ends. See also Instances Overview.	Instance Availability Management
Remove Schedule	Cancel an existing stop/hibernate schedule for selected instance(s). See also Instances Overview.	Instance Availability Management
Notification Settings	Configures email notifications for selected lifecycle events of one or more instances. See also How to Configure Instance Availability Notifications. Supported events:: an instance is hibernated an instance is woken up a downtime schedule is created a downtime schedule is changed a downtime schedule is removed	Instance Availability Management

Limitations for Bare Metal Instances

Bare metal instances do not support instance availability actions (Hibernate, Wake Up, Set Stop Schedule, Remove Schedule ) that apply only to cloud deployments.

These actions do not appear for bare metal instances in the Instances view.

Availability Change Failure Alerts

When PlatformManager cannot complete an availability action (hibernate, wake up, create or modify a schedule), it performs the following:

Displays a failure notification in the PlatformManager Notification Center
Sends an email to the user who initiated the action
The email contains: the name of the instance, the action that failed and the returned error message.

This ensures that the requesting user and the PlatformManager team can react immediately without waiting for a support ticket.