Replies: 9 comments 1 reply
-
|
Hi @tep1997, sorry for the friction on this one. Short answer: 1.4.0 is a clean install on top of a fresh Postgres database. There is no in-place migration from a 1.3.x SQLite install. Why we went this way. We seriously considered keeping both backends side by side (Prisma talking to SQLite and Postgres) but carrying two database adapters means doubling every Prisma migration, every raw query we tune for vDC quotas / IPAM / reports, every test fixture, every CI matrix. SQLite was already where we hit lock contention as soon as multi-tenant workloads, vDC quotas, replication metadata and the report generator started running concurrently. For where the product is heading next (HA via Postgres streaming replication, a shared cluster store between frontend and orchestrator, cross-tenant JSONB aggregations), Postgres is the right substrate. A maintained SQLite branch on top of that would slow every future release for everyone, including you. Recommended upgrade path:
What you do not need to redo: the VMs and CTs themselves live on Proxmox, so the inventory scan picks them up automatically once a connection is re-added. Backups on PBS are unaffected too, you just re-bind the PBS connection. What you do lose: audit history, in-app task / event history, in-app alert rules. If audit retention matters for you, export it from the 1.3.x UI before stopping the stack. We'll add a proper "Upgrading from 1.3.x" section to the repo so this isn't only in a discussion thread. Thanks for flagging it. |
Beta Was this translation helpful? Give feedback.
-
|
Ok, makes sense. Thanks for the reply. I used a docker compose file to install originally if I'm not mistaken. Is that still an option? |
Beta Was this translation helpful? Give feedback.
-
|
Yep, compose-only is still fully supported, the
One thing to watch: the old 1.3.x compose pointed at a SQLite file under (Edit: fixed the installer URL, it's |
Beta Was this translation helpful? Give feedback.
-
|
Ok, after carefully considering your post, I decided to scrap my current setup and just start from scratch using the script. The curl command wouldn't work for some reason (could not resolve host error). I manually copied the script and now have a working instance of the new architecture. Thanks for all the work! |
Beta Was this translation helpful? Give feedback.
-
|
Glad you got it running, and thanks for sticking with it! Quick follow-up on the |
Beta Was this translation helpful? Give feedback.
-
|
This is what I ran: curl -fsSL https://get.proxcenter.io/community | sudo bash Error:
I ran it on multiple hosts with same results. Looks like perhaps the subdomain doesn't resolve, but the root does:
|
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the screenshot, that nails it. The Two of the install scripts still carried the wrong URL in their header comment / upgrade hint ( |
Beta Was this translation helpful? Give feedback.
-
|
Finally, what is the proper way to update a non-docker install that was setup using the script? |
Beta Was this translation helpful? Give feedback.
-
|
The install script you ran is actually a Docker install under the hood. It installs Docker Engine if it isn't already there, writes You can confirm with: You should see That re-pulls If on the other hand you genuinely have a non-Docker setup (Node running directly under systemd against a native Postgres), that isn't a path we ship today. The only supported install is the Docker one. |
Beta Was this translation helpful? Give feedback.


Uh oh!
There was an error while loading. Please reload this page.
-
What is the recommended upgrade path to convert to the new 1.4.0 release? The standard update commands break the frontend as you mentioned in the update release. How then do we upgrade, or do we start from scratch?
Beta Was this translation helpful? Give feedback.
All reactions