Changelog

v26.1.0.2 - 2026-09-03

  • Upload the whole backup archive on every attempt of a ProvisioningRestore. The archive was passed to the request as an open file, and a network function that is still warming up answers 502/503, which the session retries by repeating the request with the same arguments – by then the file was at its end, so the retry uploaded an EMPTY archive. The api accepts that upload and the restore then terminates with cannot_decrypt, which before the outcome check below was reported as a successful deployment: an upgrade that restored none of the provisioned data. The file is now re-read per attempt instead of being held in memory for the whole call

  • Ride out the network-function warm-up window on a provisioning restore: retry a 503 on every request (a network function that is still starting behind the athonetos gateway answers 503) and, for the restore request only, retry a 404 for up to five minutes (the gateway answers 404 for the path of an application that is not mounted yet, which is the case for a while after a vm boots or is re-imaged). A 404 stays fatal everywhere else, as the handlers rely on it to detect an absent resource

  • Fail the deployment of a ProvisioningBackup/ProvisioningRestore when the api reports that the action terminated without succeeding. The action status only ever takes the values CREATED, RUNNING and TERMINATED, and whether the action achieved anything is reported separately in info.outcome, so waiting for TERMINATED alone reported a restore that restored nothing as successfully deployed. Warnings the api reports about the action are logged

  • Retry the base-config PUT in the test reset fixture through the post-start write window: a just-started service (observed on the eir service of the 24.3.1 image) answers GET .../mgmt/config with a 200 a moment before it accepts writes, intermittently failing the reset with a transient 400

v26.1.0.1 - 2026-06-10

  • Only emit the CHF diameter block on athonetos 24.3 and later: older versions have no CHF diameter support (24.2.2 rejects a configuration that carries it), while from 24.3 on the device crashes when the block is absent

v26.1.0 - 2026-06-03

  • Add support for athonetos 26.1.0 (regenerate the module against the 26.1.0 API)

  • Strip the PCF policy script type field on athonetos 24.x (PCF API 1.2.x), which rejects it

  • Emit an explicit empty scripts list for the PCF SM/AM policy on all versions (the serializer drops empty relations, leaving "am": {}, which the device’s set_config rejects)

  • Emit an explicit empty diameter block for the CHF config (the serializer omits the empty relation, which the device’s set_config rejects)

  • Index Api on (base_url, agent_name) instead of base_url alone, so a distinct agent_name yields a separate Api against the same box (allows a dedicated agent per resource type, e.g. slow license fetch/activation)

v23.4.2.5 - 2025-07-09

v23.4.2.4 - 2025-04-03

  • Add support for 24.3.1

  • Fix Pls configuration for 24.x

  • Make Upf update more resilient

  • Add Smsf support

  • Fix smf configuration when no dnn specified

v23.4.2.3 - 2025-02-07

  • Allow to persist pls component config

  • Add support for vm reboot after upf dataplane changes

  • Enable auto-start for all services

v23.4.2.2 - 2024-11-28

  • Fix dry-run issues on broken clustering.

  • Disallow leaving a cluster, unless specified otherwise in the desired state.

v23.4.2.1 - 2024-11-18

  • Retry clustering success check

  • Fix dry-run issues in 24.2 forward compatibility

v23.4.2.0 - 2024-11-06

  • Regenerate module to stop using deprecated number type

  • Add athonetos 24.2 forward compatibility

v23.4.1.7 - 2024-09-16

  • Fix password status url wait mechanism in case of connection error

  • Add default dnn to the smf body

  • Improve error reporting in waiter

v23.4.1.6 - 2024-07-16

  • Enable inherit_config_element feature from restbase, to allow deleting objects from lists

v23.4.1.5 - 2024-06-25

  • Restart UPF without retry on 50x error codes

v23.4.1.4 - 2024-06-03

  • Improve password handler error handling

  • Add connection and http retry attributes to api resource.

v23.4.1.3 - 2024-04-23

  • Update inmanta-module-restbase version

v23.4.1.2 - 2024-04-16

  • Disable stack endpoint for PCF component

v23.4.1.1 - 2024-04-08

  • Add reboot and waiting capabilites for generated resources

v23.4.1 - 2024-03-27

  • Add support for backup/restore mechanism

  • Add support for backup/restore of provisioning data

  • Add support 23.4.1

v23.4.0 - 2024-03-11

  • Added username to Password entity

  • Added support for 23.4

  • Added Pls component

v23.3.0 - 2024-02-15

  • Stop using deprecated jsonschema.RefResolver

  • Added Password entity to enforce a password

v23.2.0.3 - 2024-01-16

  • Added basic api co-management support.

  • Added identity information for Smf component

  • Fix logging of non-existent variable in Cluster resource

v23.2.0.2 - 2024-01-03

  • Added support for license precedence

v23.2.0.1 - 2023-11-27

  • Added support for 23.2

  • Added support for subresource APIs

v23.1.0.1 - 2023-09-26

  • Added support for 23.1

v5.0.2.1 - 2023-08-03

  • add support for netbox-secret integration

  • Added attribute soft_fail_status_codes to BaseResource to allow skipping resources for certain status codes.

  • Improve dry-run checks

v0.1.0 - ?

  • Add retry capabilities to the api handler

  • Initial release