Hi all,
For various reasons, I’ve reached the limits of NodeRed on the Large CerboGX image and need to offload to a more powerful box. I need to be able to use the Victron nodes and a bit of searching seems to imply that this is possible, but… I gather I need to make the dbus information available to the remote NodeRed server which is a two-step process - firstly get the Cerbo to make it available, and secondly get the remote NodeRed server to access it. Unfortunately, I haven’t been able to find any instructions for exactly how to do this - my search-fu is broken!
Does anybody have any links to documentation which detail the steps required?
Cheers,
N.
I would recommend using MQTT.
Use the Victron nodes to publish the relevant data points to a broker (any broker, I don’t use the Cerbo’s broker) and your remote node red instance subscribes.
Your remote node red can then do whatever it needs to and publish relevant commands via the broker. Your Cerbo subscribes and uses the Victron nodes to effect the commands or changes
1 Like
Hi Deon, thank you for your comment.
Unfortunately that isn’t a direction that makes sense for my particular problem - I would need two NodeRed servers with logic in both of them. The one on the cerbo would publish the Victron settings to MQTT and subscribe to the parsed/processed replies from the new NodeRed.
It would be far far easier/faster/better to run just one NodeRed with the Victron nodes and do everything from there. Even if this were for no reason other than being able to backup the entire configuration.
N.
Phew, a bit more searching and I’ve got it sussed, tested and working…
This might not be an optimal solution, It is possible to run the entire CerboGX operating system on a Raspberry Pi and I believe modern Raspberry pi out perform CerboGX in relation to CPU an memory. But this might be introducing new challenges by doing this. There are videos on Youtube about this.
I am curios to know, what do you mean by “reached the limits”. Was there a noticeable performance problem on the CerboGX? Do you think the CerboGX Memory or CPU was under too much load?
I’m a software engineer with prior experience with the Linux OS which the CerboGX is running on a variant of Linux.
However I am also having problem with NodeRed on my CerboGX (actually I have a Multiplus 2 GX which has similar capabilities to a CerboGX)
Hi, @expertcoder, thank you for your reply.
I have a RPi running VenusOS in my caravan (travel trailer) and a Cerbo in my campervan.
I’m not going to go into the “reached the limits” bit - it’s not really relevant and I don’t want to have to deal with a whole hoard of people assuming my lack of competence and telling me that everything would be fine if only I were to do X, Y or… All I will say is that I can have a working configuration on a RPi that breaks on a Cerbo (again, not going to go into the details). Just trust me!
I already have an Intel NUC running Proxmox (with containers and VMs running RouterOS, Home Assistant, Mosquitto, Zigbee2MQTT, influxdb and Grafana), so to me it made sense to move the NodeRed functionality off the Cerbo and onto that box - if nothing else, it would mean that NodeRed gets backed up to my Proxmox Backup Server.
Having now built a NodeRed container, installed the Victron nodes and enabled external dbus access on the Cerbo, it just works. Sure, there’s a bit of functionality missing, but everything I want access to, I have access to.
Thank for the reply, you seem to be running lots of interesting hardware. I know why you would not want to run NodeRed on the CerboGX when you already have substantially more powerful hardware in the same Network.
I am having an issue with my NodeRed running on my Victron Multiplus 2 GX, where the NodeRed service on the underlying OS is been stopped (I dont think it is crashing, something is instructing it to stop).
I don’t think my NodeRed flow is resource hungry, but I think the problem only started after I made my NodeRed flow bigger.
You could consider installing the Home Assistant node red addon and bridging the cerbo to the home Assistant mqtt broker. I have done exactly that and my victron ecosystem is fully integrated and controlled by my HAOS on a raspi 5.
Hi, @thomasinaz, thank you for your reply.
Unfortunately it doesn’t answer my question or address my situation. However, I’m happy you have a solution that works for you.
N.
I was simply suggesting that you leverage the resources that you are already using. I would be happy to provide details on how to do that if you are interested.
Hi @thomasinaz
In the spirit of this board, I was trying to be polite. Put a little more bluntly, the solution you suggest is based on a number of incorrect assumptions and is not suitable for my use case.
I’m not going to discuss my requirements and decisions further.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.