Hi pioreactor developer and user,
I am new comer to Pioreactor, but have manged to run it as a turbidostat with E. coli culture for 50+ hours. It went smoothly except for two abnormalities.
While the UI was normal when initiating the experiment, when re-open UI say 1 hour afterwards, its Pioreactor page showed that all functions of the device were off even though they were not and the Overview page was churning out latest readings.
This problem does not do harm to my experiment but the second problem led to the pre-mature termination of the experiment:
After 50 hours (the experiment were continuously running), when re-opening UI, it started to prompt “failed to connect to MQTT. Is configuration for mqtt_broker_address correct? Currently set to synevol.local”. I cannot control the device by the UI. I tried to change the mqtt broker address in Configuration file to that suggested in another post. I did not work.
I know this problem. How are the pioreactor connected to you network? I solved this by connecting the pioreactors, each one via LAN to a LAN-Switch and my laptop which i am using to navigate the pioreactors also to the switch. This way you can’t loose the connection to MQTT broker. But sadly this also means to not be able to work with them without LAN-connection and standing right next to them. For me this is the case because the institution i am working in has a crazy firewall and won’t “accept” the MQTT broker (or something like this). Maybe connection via Ethernet can also save your problem. I am also trying to find a fix for this problem, so if anyone has an idea about the MQTT broker let me know
Thanks a lot for your fast response, Leo!
I was using public wifi of the building which does not have a crazy firewall. Plus, why the first 50 hours was fine? I am puzzled.
If wifi is the problem, how about physically visiting the device and turning on hotspot of the lactop for Pioreactor to log in each time when we want to check the UI? When we are away, the data are continuously deposited to the local SD card.
Would this work as long as keeping the experiment going?
@LeoKleym I don’t know the specific of your institutions firewall, but you’ll need to atleast allow ports 80, 9001, and 1883 to accept connections for the Pioreactor. Maybe this is what the IT team needs?
While the UI was normal when initiating the experiment, when re-open UI say 1 hour afterwards, its Pioreactor page showed that all functions of the device were off even though they were not and the Overview page was churning out latest readings.
Hm, did a refresh of the page “correct” this. I wonder if there was a lost connection to MQTT (which publishes this live data), and there wasn’t any reconnection after it was lost. I’ll look into adding better reconnection attempts.
After 50 hours (the experiment were continuously running), when re-opening UI, it started to prompt “failed to connect to MQTT. Is configuration for mqtt_broker_address correct? Currently set to synevol.local”. I cannot control the device by the UI. I tried to change the mqtt broker address in Configuration file to that suggested in another post. I did not work.
Ah, that’s frustrating - sorry about this! Sometimes, a page refresh would solve this, but it should like it didn’t in your case.
In either case, if the MQTT broker was still up on the Pioreactor (where it resides), and the problems you are seeing are just on the client (your browser), the Pioreactor should still be recording data.
One thing to check is the Pioreactor logs. You can SSH into your Pioreactor and check pio logs for any MQTT errors.
Hi @CamDavidsonPilon
Refreshing or re-opening UI did not help with either problem.
Is it possible to circumvent wifi by connecting Pioreactor with a laptop through a cable, and operating and/or monitoring Pioreactor with the laptop?
Or, can I turn on wifi hotspot from my cell phone to which both a laptop and Pioreactor are connected? When initiating the experiment or checking ongoing runs, I simply bring the laptop and cell phone with wifi on to the Pioreactor.
