We have both MQTT and Cedalo running within a docker environment in Debian. System has 2 cores (CPU) and 16GB of RAM as we noticed memory utilization being quite high.
Recently over the weekend it appears that any time you log in to the Cedalo webpage I’m met with the image/screen “We could not connect to your broker”. This prevents any interaction with the webpage entirely, so even checking connections or configurations isn’t possible via web.
This also goes hand in hand in noticing that CPU/Memory will spike, hang the system and the container for Cedalo ends up crashing / restarting. To my understanding Cedalo is simply a front end to display the data that the broker (MQTT) ingests; So I would not understand why Cedalo would or could be the reason for such resource intensivity.
I’ve deleted and recreated the container for Cedalo a few times now with fresh configurations and appears that the crashing stabilized for the most part, though this morning I observed again resources spiking high, causing a crash and restarting the Cedalo container.
Accessing the Cedalo home page on the actual VM that it is running on allowed me to access without any issue. Top right corner I could see Mosquitto (the broker) as a valid connection and no real issues past that. Today I’m experiencing the same “We could not connect to your broker” on both the local system as well as attempting to access the webpage remotely.
All this is to say, the system statistics in the background are all updating and displaying accurate information on the data ingested etc through MQTT. Realistically everything looks to be functioning, just actual web ui access is a dud.
Can you tell me exactly what you are using? We have two products, Mosquitto and Streamsheets.
The Mosquitto broker has a Management Center as a web UI.
Both are either open source or as a paid version available.
Could you clarify what product, version and if paid or OS you are using.
As I was attempting to gather that screenshot, CPU util shot through the roof, and the Cedalo instance in the docker container had restarted. This is a somewhat reoccurring issue now, I’m wondering if there is a limitation on the number of data ingested, could we be exceeding it?
Out of interest, I accessed the management center from my laptop (remote) and monitored CPU utilization as well as memory utilization.
Upon logging in, it definitely spiked quite a bit from low 1% to nearly 60% peak before settling at around 50% util. Refreshing caused a spike before once again settling.
While just refreshing and seeing its effect on utilization, it now suddenly “works”, which will be short lived before we are back to the previous image above.
