Benefit of logging Apex data independently

[gbru316]

I noticed earlier today that my Apex (A2) had stopped logging data --- pH, ORP, Temp, Alk, Mg, Ca.

After going through the typical recycle power - factory reset - reload config path, I ended up calling Neptune when I also noticed that I could not change the configuration. For instance, I was unable to change the COR-15 speed via fusion. I noticed that each time I tried to view the configuration page for the pump, it prompted that I had not yet run the setup wizard despite just doing so. Same for Trident controlled dosing.

After exhausting my options, I called Neptune. They confirmed that there's a faulty component in the controller and are processing a replacement. I asked if they could tell whether dosing was occurring or not and unfortunately, they couldn't.

BUT, then I remembered that I had set up a TiG stack on an old linux PC I use for various other hobbies. When I viewed the graphs, I was surprised to see that after I restarted the Apex, it was indeed feeding data that was missing from Fusion into the dabase. Though, I wasn't sure how accurate it was given the issues with the Apex. So I ran a quick manual Alk check and it was within the expected margin of error. Which means, Trident is still measuring and dosing, it's just not recording. My tank is predominantly SPS and it appears like the issue with the Apex started 24 hours ago, so if it wasn't dosing I'd expect Alk to be significantly low. It's not.

As you can see in the graph, logging stops at 10 am yesterday. I restarted the Apex between 10:30 and 11 this morning. The red oval outlines the period for which Apex wasn't logging (OR measuring). But after the reboot, it's at least measuring.

Here's 9 months of alk data:

Had I not set up the external database, I'd be stuck manually testing and dosing for the next 24-48 hours (or however long the cross-ship takes). This provides some peace of mind and reduces some of the (unexpected) work of keeping a reef tank. Which is why we buy controllers in the first place, right?


Anyone interested in setting up a similar database can follow the instructions/thread on the neptune forums.

 

[areefer01]

As a data hoarder I will say that I also poll and collect data from my Apex. Telegraf, influxDB, Grafana. Once you have the data you can do all sorts of stuff.

 

[OfficeReefer]

I suspect this occurs more often than not. Their concept was originally designed with power and literally a "controller" pad on the outside that one would do things on. The next versions were adapted with the multi-tenant and cloud panel to access and scale.

It would be best if this all logged locally and then was sent to the cloud, however I would be surprised if they aren't cutting costs like everyone. This includes storage, network logging and streams, etc.

I wish there was a company that made pool controllers and aquarium controllers...

 

[A_Blind_Reefer]

I think this memory failure is a bigger problem than we think. What I found when it happened to me a month or so ago was that my wav pumps and dos units continued to operate. My conclusion was that those devices must store and run their program locally at the device. Everything else was locked in whatever state it was in when the memory failed. So return pump, skimmer, co2 solenoid, lights, etc we’re stuck on 24/7. Things like feed modes, heaters, chillers, etc were stuck off 24/7. Along with that was the deletion of all programming, including internal and external alarm notifications. So, without randomly logging into fusion and noticing the missing data, I can’t see to look at the graphs so I didn’t notice right away, you wouldn’t know….. unless you’re by your tank 24/7 and notice the lights staying on or something. I’ve been asking everyone that has had this failure to ask Neptune for an updated heartbeat notification in fusion to notify when there is a memory failure. I would think it wouldn’t be too difficult as the apex still communicates with fusion. Maybe a built in monitored virtual outlet that switched on and off every hour, that if it doesn’t switch a notification is sent out. I ended up hooking something up on my own for this. If enough of us complain maybe they’ll do something. Someone could lose everything if they weren’t paying attention for a while….. which is part of the point of having an apex.