commit 814d63735016935ba0ca0bbc9ae277067341a03d Author: Sorin Savu Date: Fri Jan 30 14:24:44 2026 +0200 30-01-2026 diff --git a/work-notes.md b/work-notes.md new file mode 100644 index 0000000..b3c633e --- /dev/null +++ b/work-notes.md @@ -0,0 +1,145 @@ +# pending defects notes: + +## Working on + +### MS-8868 -The OS is not reinstalled when the templated id is changed on the server instance group + +***TODO:*** +Reproduce this and fix. I have two requests to look into this from Alex Bordei and https://chat.metalsoft.io/bigstep/pl/751ugq9xrtg8tkgaz7pz1xs1xc +Sergiu ? + + +### MS-9347 - Unhandled exception causes chain restarts in the Inventory ms and erratic behavior for the server after multiple deploys + +**Multiple actual issues:** +* inventory microservice was exiting on a SQL error due await missing - added await +* the size of the field for the agent_version was limited in the database - updated the size to 255 for both the `agent_version` and the calculated field `agent_type_and_version` +* there was a panic in the ms-tunnel, handled both by me and Sergiu - ***TODO*** - verify how to reproduce +* updated the version checking in the ms-tunnel to convert known tags to a semantic version - ***TODO*** - clarify the changes, and document them + +**TODO** +* add changes summary and test hints in the defect +* reasign to Andrei + + + + +### MS-8510 - Zombie Ansible processes + +1. TODO: Try to reproduce this +2. Remember that there seems to be an issue with how the config is located, check more +3. + + +## Waiting for + +### MS-8883 - Update ms-agent to support cumulus-linux ZTP + +~~Asked Eduard and Ciprian for status~~ + +ZTP flow expects that the switch exists in the database. We need to change that so it's creating a switch if missing. TBD how + +**Meeting notes:** +* According to Sergiu this is a known limitation +* It should be possible to use the same AFC flow as for ONIE +* There doesn't seem to be a defect/story for it +* Unknown priorty, check with Mike if needed +* should be able to test in GNS3 +* MS-8883 can't be tested because they don't have a image compatible with GNS3 + + + +### MS-9358 - Remove syslog priority filter from API and backend + +Sent to Andrei on 30 jan 2026, awaiting response, will check next week + +Solution: I just removed the fields from the ms-tunnel/ms-agent structs and removed code that was procesing priority stuf + +## General defects + +### MS-3684 - Make SSH password protected keys be usable when uploading to the firmware binary repository + +1. TBD: how do we store passwords for keys +2. POC: check how to use the existing SSH library to connect using password protected key, and check if there is support for a ssh agent + +### MS-5288 - Implement INBAND DHCP agent in the ms-agent Golang project and the tunnel + +1. Onboarding: are there examples of in-band networks connected to agents in the existing setups +2. POC: search for a json/record based database that is light weight and can be integrated in the agent +3. POC: try sqlite for this + +Sergiu ?I need to figure out what exactly is the inband-outband difference + +### MS-5743 - move the factory reset sequence into a driver specific function + +code is still in the controler, it should be switched to a specific driver, but it has a guard now. Maybe close it silently + +### MS-5744 - generalize implementation to make it clear it works for both servers and switches + +This is about the kafka event used for SSH execution, that should work for switches also. They have repurposed it, I should check that out. +Also, I was thinking that some checks/loops should be moved in the agent, and not wait actively upstream + +1. TODO: review the existing code to figure out the best solution + + + +### MS-5994 - Add support for syslog collection for the incus app hosts + +1. TODO: find a way to figure out why is this assigned to me +2. TODO: check the API for incus for configuring the syslog +3. TODO/onboarding: learn what is the implementation for incus + +### MS-6251- OS10 switch ZTP implementation + +1. TODO: review the status of this feature, how is integrated with switch defaults +2. this is moslty done, but I don't know if it was tested, see who can/if was tested + + +### MS-6896 - Refactor to capture errors before confirming agent registration + +1. TBD: this should be part of the big rewrite, better error handling here is crucial. +2. TBD: I need to do some research on the life cycle of the tunnel/agent system, and design shutdown cycle + +### MS-7443 - Reimplement Fallback for OS Template Selection via Template Microservice + +1. review the current implementation, since it seems that it was changed/fixed +2. Sergiu ? + +### MS-7455 - Implement support for HTTPS Proxy tunneling in ms-agent + +1. TBD: is this done ? seems so +2. sillently close it, after confirming everything works + +### MS-7634 - [GraphQL] TunnelConfig type should not be editable, it should always be TunnelConfig + +1. TBD: review what Alex implemented and check if this has went away + +### MS-7918 - Add support for DHCPv6 server registration + +1. TBD: what are we doing here + + + + +## Extension migration defects + +### MS-7407 - Migrate extensions microservice to the nodejs-metalcloud repo + +1. TODO: Review the tests and make sure the coverage is complete +2. TODO: Start moving validation from manual to class validator + +### MS-8447 - Allow update of workflow extensions + +1. TBD: clarify what the fuck is the requirement here + +### MS-8475 - Add support for extension versioning + +1. TBD: clarify what the fuck is the requirement here + +## Morpheus + +12:27 PM + +https://morpheus.metalsoft.dev +admin +4xfC3B6LCMMlRiuQmA2kWWXo3L8=