These are chat archives for nightscout/intend-to-bolus

14th
Feb 2019
Dave Gourley
@RadEnder
Feb 14 04:10
Screen Shot 2019-02-13 at 9.08.52 PM.png
Having a new issue with Nightscout going down. I'm getting notifications from UptimeRobot saying that my site is going up and down. When I try to login to my site, I'm asked for my API secret but am then told that it is incorrect. I've retyped it multiple times just to be sure I wasn't fat fingering and I'm confident that I'm not. Papertrail is showing my app is crashing as shown. I did see that @garykidd may have experienced this same issue and may have fixed it by deleting his logs but I'd rather not do that if I don't have to. I did try to redeploy NS but it that didn't work either. Any ideas?
Dave Gourley
@RadEnder
Feb 14 07:33
For anybody else having the issue I described above it did turn out to be my mlab database size which had reached its limit. Once I cleared out the devicestatus logs the site returned to normal. I didn't back up the database because I just wanted to get back online so I'm hoping that was the right move. We're up again at least! Happy Looping!
tuzoenduro
@tuzoenduro
Feb 14 09:52
hi @RadEnder , yeah, as I have seen, whenever NS starts refusing the API secret it's time to clean mlab...
Jan Schenk
@jansche
Feb 14 10:04
Has anyone experienced the "Rebooting to fix radio errors!" problem? Ran into it with an Edison rig today. What's the background?
peterfulvi
@peterfulvi
Feb 14 15:25
@ecc1 Ok i got past the ID proccess but now my pump is not responding
root@LGRIG:~/myopenaps# pumphistory -n 6
2019/02/14 10:23:13 retrieving pump history since 2019-02-14 04:23:13
2019/02/14 10:23:14 connected to CC111x radio on /dev/spidev5.1
2019/02/14 10:23:14 setting frequency to 916.600
2019/02/14 10:23:15 waking pump
null
root@LGRIG:~/myopenaps# export MEDTRONIC_PUMP_ID=xxxxxx root@LGRIG:~/myopenaps# killall -g oref0-pump-loop; pumphistory -n 6
Eric
@ecc1
Feb 14 15:37
Can you try running some mdt commands, like mdt model and mdt reservoir, and try pumphistory for a shorter time ("-n 1" for one hour say)
peterfulvi
@peterfulvi
Feb 14 16:06

@ecc1 I ran sudo service cron stop
Then

root@LGRIG:~/myopenaps# mdt model
2019/02/14 11:03:08 connected to CC111x radio on /dev/spidev5.1
2019/02/14 11:03:08 setting frequency to 916.600
2019/02/14 11:03:10 waking pump
2019/02/14 11:03:25 no response to wakeup

Then

root@LGRIG:~/myopenaps#  mdt status
2019/02/14 11:06:01 connected to CC111x radio on /dev/spidev5.1
2019/02/14 11:06:01 setting frequency to 916.600
2019/02/14 11:06:03 waking pump
2019/02/14 11:06:18 no response to wakeup

No response from pump

The Pump is next to me, also its a serial connection
:confused:
Eric
@ecc1
Feb 14 16:16
What's a serial connection? This is an Explorer Board, right? And you've set the pump serial number correctly in the MEDTRONIC_PUMP_ID variable?
If so, try running mmtune; that default frequency might be way off for your pump
You can also set the MEDTRONIC_FREQUENCY environment variable yourself if you suspect it should be something else. This is a North America model pump, right?
(Run mmtune -ww if it's a "worldwide"/European pump)
peterfulvi
@peterfulvi
Feb 14 17:17
@ecc1 The serial connection is connected by usb cable. Its an Edison Ilooked in oref0-runagain.sh and the serial number is correct. I think its a US pump PARxxxxxx. Thats wht i set it up. Far as mmtune -ww root@LGRIG:~/myopenaps# mmtune -ww -bash: mmtune: command not found
I ran mmtune -us for the heck of it and same thing
Scott Leibrand
@scottleibrand
Feb 14 17:20
if you have the letters PAR in the S/N, remove those to leave just the numerals
if you don't have mmtune installed, you might not have oref0 dev installed correctly...
peterfulvi
@peterfulvi
Feb 14 17:24
Ok. I dont have PAR in the numbers. So do I have to run setup again or can i jus update the dev branc?
Scott Leibrand
@scottleibrand
Feb 14 17:25
I would do a fresh cd ~/src/oref0 && git checkout dev && git pull, then npm run global-install, check if it's working, and if not re-run oref0-setup
peterfulvi
@peterfulvi
Feb 14 17:44
@scottleibrand Thanks
Eric
@ecc1
Feb 14 17:48
@peterfulvi the NA/WW distinction is encoded after the model number on the pump's label. For example, MMT-522NAB is a North American (Blue) pump.
And it looks like the setup script installs mmtune as Go-mmtune
Scott Leibrand
@scottleibrand
Feb 14 17:58
it also sets up a symlink/alias so mmtune works
peterfulvi
@peterfulvi
Feb 14 18:20
@ecc1 Its NAS
North American States? :smirk:
Eric
@ecc1
Feb 14 18:31
I think the S is for smoke (gray) color
peterfulvi
@peterfulvi
Feb 14 18:37
oops. My bad. I updated and still no mmtune. I'll rerun setup later and let you know. Thanks :thumbsup:
Mike
@libxmike
Feb 14 19:29
@scottleibrand @danamlewis still not sure what time/date range take to run autotune to adjust my basal rate, any recommendation?
Scott Leibrand
@scottleibrand
Feb 14 19:29
I would recommend running it over all available data
Mike
@libxmike
Feb 14 19:38
thats 2 years :)
OpenAPS and AAPS
Mike
@libxmike
Feb 14 19:46
will try to run it overnight
Scott Leibrand
@scottleibrand
Feb 14 19:57
hah, ok. probably diminishing returns beyond the last month
Jan Schenk
@jansche
Feb 14 20:46
Any hints on why a running and looping rig doesn't show up in the nightscout openaps pill?
Scott Leibrand
@scottleibrand
Feb 14 20:47
check /var/log/openaps/ns-loop.log on the rig
Mike
@libxmike
Feb 14 21:16
@scottleibrand so 30 days should be enough?
@jansche also could be mLab DB maxed out (if you using it) if this is the case, delete all entries in devicestatus and go to tools - commands and repair DB
Jan Schenk
@jansche
Feb 14 21:28
@libxmike other rigs work fine, so it's probably not mLab? Problem came in after updating...
Mike
@libxmike
Feb 14 21:30
check logs which Scott posted, maybe you find error there
but still check mLab
Jan Schenk
@jansche
Feb 14 21:36
@scottleibrand : hah, thanks, unauthorized. Copied the Api key with one digit missing.
Where do I change the NS API key without rerunning setup?
Thanks for the pointers, @libxmike and @scottleibrand!
Mike
@libxmike
Feb 14 21:40
ns.ini and crontab, if I remember correct
Jan Schenk
@jansche
Feb 14 21:40
@mkericson that's what I want to avoid. It's a RPi zero. Takes an eternity...
Jon Cluck
@cluckj
Feb 14 21:41
ns.ini, cron, preferences.json, and .bash_profile
Jan Schenk
@jansche
Feb 14 21:42
It's hashed in there though.
Never mind. I know there are ways to hash it on the command line, but I was just too lazy to find out how exactly. Ran runagain.
Jon Cluck
@cluckj
Feb 14 21:57
:laughing: I didn't suggest how to do it because I too don't remember and am lazy