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

18th
Oct 2017
garykidd
@garykidd
Oct 18 2017 00:00 UTC
@scottleibrand why would MinGuardBG go to 999?
dmdfreak
@dmdfreak
Oct 18 2017 00:02 UTC
I have zero carbs in my system and zero bolus or smb insulin in my system. But for some reason it's dropping me into the 60s even with my target raised to 92-94. I'm on Dev and updated this afternoon. What could be causing it to take me low?
Scott Leibrand
@scottleibrand
Oct 18 2017 00:28 UTC
@garykidd that's the initial value it's set at before getting lowered for predBGs. It basically means it couldn't calculate minGuardBG for some reason. I'd have to see logs to guess why.
@dmdfreak "it's dropping you" by administering extra insulin? or "you're dropping" and it hasn't been able to counteract? I'd need more details (maybe NS screenshot) to see what's going on
dmdfreak
@dmdfreak
Oct 18 2017 00:29 UTC
I'm dropping. I'll post a screen after dinner.
garykidd
@garykidd
Oct 18 2017 00:45 UTC
@scottleibrand FWIW...still seeing instances of manual boluses not being depicted in NS...similar to previous reporting...
KS615
@KS615
Oct 18 2017 00:54 UTC

Hi all, I'm trying to clean my mlab and I keep getting the following error, ""message": [
"The repairDatabase command was issued, but the request timed out.",
"Usually this means the command was successfully started in the database,",
"but it will take a while to complete."

Hoping this is an easy answer that I just can't seem to figure out. I was following the instructions in the docs.

Scott Leibrand
@scottleibrand
Oct 18 2017 01:14 UTC
@garykidd so far no one is working on fixing that issue (that I know of), just gathering data
garykidd
@garykidd
Oct 18 2017 01:53 UTC
@scottleibrand cool. Let me know if I can help in any way. Keep on rockin in a free work. RIP TP.
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:19 UTC
I tried installing the newest dev tonight, but I'm getting errors now that are preventing the rig from looping:
Oct 17 22:07:02 natopenaps pump-loop.log:  Starting basal-only pump-loop at Tue Oct 17 22:07:02 EDT 2017 with 8 second wait_for_silence:
Oct 17 22:07:02 natopenaps pump-loop.log:  grep: enact/smb-suggested.json: No such file or directory
Oct 17 22:07:02 natopenaps pump-loop.log:  Waiting up to 4 minutes for new BG: ls: cannot access /tmp/pump_loop_completed: No such file or directory
Oct 17 22:07:13 natopenaps pump-loop.log:  Radio ok. Listening: .No interfering pump comms detected from other rigs (this is a good thing!)
Oct 17 22:07:38 natopenaps pump-loop.log:  Preflight OK. Profile less than 60m old. Refreshed pumphistory and meal.json
Oct 17 22:07:40 natopenaps pump-loop.log:  Checking pump clock: "2017-10-17T22:07:31-04:00" is within 1m of current time: Tue Oct 17 22:07:40 EDT 2017
Oct 17 22:07:45 natopenaps pump-loop.log:  and that pumphistory is less than 1m old.  Temp refreshed
Oct 17 22:07:47 natopenaps pump-loop.log:  usage: openaps-report invoke [-h]
Oct 17 22:07:47 natopenaps pump-loop.log:                               {cgm/cgm-glucose.json,cgm/ns-glucose.json,enact/enacted.json,enact/suggested.json,monitor/battery.json,monitor/carbhistory.json,monitor/cgm-glucose.json,monitor/clock-zoned.json,monitor/clock.json,monitor/iob.json,monitor/meal.json,monitor/mmtune.json,monitor/pumphistory-zoned.json,monitor/pumphistory.json,monitor/reservoir.json,monitor/status.json,monitor/temp_basal.json,raw-cgm/raw-entries.json,settings/autosens.json,settings/basal_profile.json,settings/bg_targets.json,settings/bg_targets_raw.json,settings/carb_ratios.json,settings/insulin_sensitivities.json,settings/insulin_sensitivities_raw.json,settings/model.json,settings/profile.json,settings/pumphistory-24h-zoned.json,settings/pumphistory-24h.json,settings/pumpprofile.json,settings/settings.json,settings/temptargets.json,upload/pebble.json,xdrip/glucose.json}
Oct 17 22:07:47 natopenaps pump-loop.log:                               [{cgm/cgm-glucose.json,cgm/ns-glucose.json,enact/enacted.json,enact/suggested.json,monitor/battery.json,monitor/carbhistory.json,monitor/cgm-glucose.json,monitor/clock-zoned.json,monitor/clock.json,monitor/iob.json,monitor/meal.json,monitor/mmtune.json,monitor/pumphistory-zoned.json,monitor/pumphistory.json,monitor/reservoir.json,monitor/status.json,monitor/temp_basal.json,raw-cgm/raw-entries.json,settings/autosens.json,settings/basal_profile.json,settings/bg_targets.json,settings/bg_targets_raw.json,settings/carb_ratios.json,settings/insulin_sensitivities.json,settings/insulin_sensitivities_raw.json,settings/model.json,settings/profile.json,settings/pumphistory-24h-zoned.json,settings/pumphistory-24h.json,settings/pumpprofile.json,settings/settings.json,settings/temptargets.json,upload/pebble.json,xdrip/glucose.json} ...]
Oct 17 22:07:47 natopenaps pump-loop.log:  openaps-report invoke: error: argument report: invalid choice: 'enact/smb-suggested.json' (choose from u'cgm/cgm-glucose.json', u'cgm/ns-glucose.json', u'enact/enacted.json', u'enact/suggested.json', u'monitor/battery.json', u'monitor/carbhistory.json', u'monitor/cgm-glucose.json', u'monitor/clock-zoned.json', u'monitor/clock.json', u'monitor/iob.json', u'monitor/meal.json', u'monitor/mmtune.json', u'monitor/pumphistory-zoned.json', u'monitor/pumphistory.json', u'monitor/reservoir.json', u'monitor/status.json', u'monitor/temp_basal.json', u'raw-cgm/raw-entries.json', u'settings/autosens.json', u'settings/basal_profile.json', u'settings/bg_targets.json', u'settings/bg_targets_raw.json', u'settings/carb_ratios.json', u'settings/insulin_sensitivities.json', u'settings/insulin_sensitivities_raw.json', u'settings/model.json', u'settings/profile.json', u'settings/pumphistory-24h-zoned.json', u'settings/pumphistory-24h.json', u'settings/pumpprofile.json', u'settings/settings.json', u'settings/temptargets.json', u'upload/pebble.json', u'xdrip/glucose.json')
Oct 17 22:07:47 natopenaps pump-loop.log:  basal-only pump-loop failed.
dmdfreak
@dmdfreak
Oct 18 2017 02:25 UTC
Settings refreshed. RefreshValueError: need more than 0 values to unpacked pumphistory and meal.json

Preflight OK. Profile less than 60m old. IndexError: bytearray index out of range

supermicrobolus pump-loop failed.

These 2 are coming up every so often lately
Brian Bonner
@itsn0tm3
Oct 18 2017 02:26 UTC
Unfortnately, looks like both dev and master are broken :(
had the same problems in dev so I decided to switch back over to master and getting this
`
Continue? y/[N] y
Checking /root/myopenaps: Traceback (most recent call last):
  File "/usr/local/bin/openaps", line 170, in <module>
    app( )
  File "/usr/local/lib/python2.7/dist-packages/openaps/cli/__init__.py", line 51, in __call__
    self.run(self.args)
  File "/usr/local/bin/openaps", line 155, in run
    builtins.dispatch(args, self)
  File "/usr/local/lib/python2.7/dist-packages/openaps/builtins.py", line 44, in dispatch
    command(args)
  File "/usr/local/lib/python2.7/dist-packages/openaps/init.py", line 6, in init
    if args.args[1] == '--nogit':
IndexError: list index out of range
Can't init /root/myopenaps
Dana Lewis
@danamlewis
Oct 18 2017 02:31 UTC
@dmdfreak bad pump read. Self resolves.
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:31 UTC
@itsn0tm3 Yes, I'm in the same boat right now. Can't run the latest dev and can't downgrade to master. I think it is a git/nogit issue.
I mean I think the inability to downgrade is related to git/nogit. I don't know what is causing the dev to fail.
dmdfreak
@dmdfreak
Oct 18 2017 02:32 UTC
@danamlewis thanks. Only been happening the last couples weeks. Thinking I better start looking for a backup
Dana Lewis
@danamlewis
Oct 18 2017 02:33 UTC
@bdr1177 yes that's a git nogit issue re master error
@dmdfreak doesn't mean an issue with the pump. Body orientation, connectivity from pump to rig, etc. nbd.
dmdfreak
@dmdfreak
Oct 18 2017 02:34 UTC
Ok cool.
Brian Bonner
@itsn0tm3
Oct 18 2017 02:34 UTC
@danamlewis how to solve then? Can't use dev unless you're using SMB since it throws an error for those missing files, but can't use master because the git/nogit option is broken?
(Been so busy I haven't gotten a chance to read up on SMB so I've been avoiding using it for now)
Scott Leibrand
@scottleibrand
Oct 18 2017 02:35 UTC
sorry, back online now
Dana Lewis
@danamlewis
Oct 18 2017 02:35 UTC
@bdr1177 I think @scottleibrand merged fixes for that in dev? On a plane about to take off so will let him pick up on suggestions s
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:36 UTC
@danamlewis and @scottleibrand Still failing with the newest fixes.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:36 UTC
let me see what else it's missing, one sec
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:36 UTC
What do I need to add to run-again to enable SMB? I can test and see if it resolves once I add that in.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:37 UTC
just 'microbolus' in with autotune etc.
Brian Bonner
@itsn0tm3
Oct 18 2017 02:37 UTC
@scottleibrand it's complaining about the 'enact/smb-suggested.json' being missing
(but would be nice to fix master too)
(sorry the first one was for dev), def haven't been a clear person all day long today ha
Scott Leibrand
@scottleibrand
Oct 18 2017 02:42 UTC
ok, I just pushed a fix for you to test
Brian Bonner
@itsn0tm3
Oct 18 2017 02:43 UTC
master or dev?
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:43 UTC
@scottleibrand That's funny because I just added 'microbolus' to see if that fixed it.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:43 UTC
cd ~/src/oref0 && git fetch && git checkout fix-pump-loop && git pull and then re-run oref0-setup
@bdr1177 that is a good test too
(or re-run oref0-runagain)
PR is openaps/oref0#729
@itsn0tm3 can you test that branch while @bdr1177 tests with microbolus?
Brian Bonner
@itsn0tm3
Oct 18 2017 02:45 UTC
ah the dev one ok sure, let me test it
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:46 UTC
Just waiting on mmtune. Why does it feel like that is the story of my life these days :smile:
Scott Leibrand
@scottleibrand
Oct 18 2017 02:46 UTC
this branch isn't merged to dev yet, so you'll need to do git checkout fix-pump-loop as I mentioned above
as soon as I get one :+1: that it's fixed I can merge it
Brian Bonner
@itsn0tm3
Oct 18 2017 02:47 UTC
hmm it says master is up to date with that branch (stratches head)
let me run that on the other rig
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:48 UTC
So with microbolus it is working. I get an unsuccessful basal-only loop and then a successful supermicrobolus loop, which I guess is expected.
Now I'll try testing the fixed branch.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:49 UTC
k. now that this rig has run with microbolus it already has the stuff I just enabled in oref0-setup for non-microbolus users
so just switching to the branch and reinstalling over the same ~/myopenaps directory won't do anything different
but you can usefully test that re-running your runagain without microbolus disables it in cron and it only runs basal-only again
or I suppose you could delete ~/myopenaps and reinstall to test it that way
Brian Bonner
@itsn0tm3
Oct 18 2017 02:51 UTC
duh me let me just delete the stuff in oref0
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:52 UTC
@scottleibrand Just to make sure I'm clear. I tested microbolus on the branch that was failing for basal-only. I didn't update with your fixed branch. You want me to test that failing branch again without microbolus?
PopcornPam
@PopcornPam_twitter
Oct 18 2017 02:52 UTC
@scottleibrand, I am reading this "Now, exercise_mode when enabled means that for a target of 111 or higher will create a manual sensitivity ratio, thus adjusting ISF and basals in proportion to this adjusted target. " and wondering if my normal target is 120, would that default it to exercise mode all the time?
Brian Bonner
@itsn0tm3
Oct 18 2017 02:53 UTC

confused look. Already on 'fix-pump-loop'

Your branch is up-to-date with 'origin/fix-pump-loop'. Already up-to-date.
there's nothing in the directory
Scott Leibrand
@scottleibrand
Oct 18 2017 02:53 UTC
@itsn0tm3 that's normal
the git fetch and checkout get you there
the git pull is just in case I push anything new
Brian Bonner
@itsn0tm3
Oct 18 2017 02:53 UTC
yeah I ran all 3 commands and the directory is empty...
Scott Leibrand
@scottleibrand
Oct 18 2017 02:53 UTC
@bdr1177 I changed my mind. let's mv ~/myopenaps/ ~/backup-myopenaps/ and then install with the new branch
@itsn0tm3 which directory?
Brian Bonner
@itsn0tm3
Oct 18 2017 02:54 UTC
~/src/oref0
excuse me while I go check my BG, my head isn't behaving the way i'm expecting it to ha
Jacob H
@jdhigh
Oct 18 2017 02:55 UTC
I'm having problems getting NS to update: https://imgur.com/a/jsEJf Any ideas? I've updated to latest NS version, packed my database (which was at 495megs in Mongo), etc. But it's been a few hours, and it's still not updating.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:55 UTC
@itsn0tm3 then let's re-clone: cd ~/src/ && rm -rf oref0 && git clone https://github.com/openaps/oref0.git
then repeat the cd ~/src/oref0 && git fetch && git checkout fix-pump-loop && git pull
@jdhigh does the DB show smaller now?
Jacob H
@jdhigh
Oct 18 2017 02:56 UTC
yes by about 200 mb
Brian Rabinovitz
@bdr1177
Oct 18 2017 02:56 UTC
@scottleibrand Okay. I just removed the myopenaps because I already backed up earlier today. I'm doing runagain with the new branch.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:56 UTC
@bdr1177 :+1: thx
@jdhigh not sure. at this point you could either try to see why oref0-ns-loop can't upload, or you could donate your data to the OpenAPS Data Commons and then take a hammer to devicestatus
I assume treatments are uploading ok?
Jacob H
@jdhigh
Oct 18 2017 02:59 UTC
yes treatments are uploading ok.
Scott Leibrand
@scottleibrand
Oct 18 2017 02:59 UTC
what oref0 version?
Jacob H
@jdhigh
Oct 18 2017 02:59 UTC
dev
Scott Leibrand
@scottleibrand
Oct 18 2017 02:59 UTC
k, then you can look at ns-loop.log and it might have something interesting
assuming you've updated dev since a few days ago when I merged oref0-ns-loop.sh
Jacob H
@jdhigh
Oct 18 2017 03:00 UTC
Yeah, last night with your bluez fix, and then again .5 hour ago.
Brian Bonner
@itsn0tm3
Oct 18 2017 03:00 UTC
back, ok the clone fixed it, running the setup now
Jacob H
@jdhigh
Oct 18 2017 03:02 UTC
find: `ns-loop.log': No such file or directory
is this bad?
Scott Leibrand
@scottleibrand
Oct 18 2017 03:02 UTC
tail -n 100 -f /var/log/openaps/ns-loop.log
oh, or just ns-looplog if you have the shortcuts installed
Jacob H
@jdhigh
Oct 18 2017 03:04 UTC
ok it's running now.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:04 UTC
it's just tailing the log, so it'll run forever. see if you can find the relevant part about ns-status uploads
Jacob H
@jdhigh
Oct 18 2017 03:06 UTC
quick question: What's that command to kick start uploading to NS?
Brian Bonner
@itsn0tm3
Oct 18 2017 03:07 UTC
okay, got one edison and one pi doing that loop now, waiting for them both to complete a loop
Scott Leibrand
@scottleibrand
Oct 18 2017 03:08 UTC
@jdhigh it should run oref0-ns-loop every minute, so you can just watch that log and it should be trying
unless you mean something different by kick-starting
Jacob H
@jdhigh
Oct 18 2017 03:09 UTC
There was some command I had to type... trying to find it in the docs. When I FIRST started, I had to type it to sort of "shock" NS to life....
No recent valid BG found. Most recent:
{"glucose":null,"dateString":null}
Refreshed temptargets: {"target":null,"duration":null}
Refreshed carbhistory; COB: 35
{"BG":174,"IOB":4.556,"rate":0,"duration":60,"units":null}
Scott Leibrand
@scottleibrand
Oct 18 2017 03:10 UTC
@jdhigh you might be thinking of openaps first-upload, but that's just for CGM entries
Brian Rabinovitz
@bdr1177
Oct 18 2017 03:10 UTC
@scottleibrand @itsn0tm3 So I have the rig successfully looping basal-only with the fix branch.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:10 UTC
:+1:
Brian Bonner
@itsn0tm3
Oct 18 2017 03:11 UTC
great! waiting for both of mine to load up the glucose file
Scott Leibrand
@scottleibrand
Oct 18 2017 03:13 UTC
I'll wait for that to finish before I merge it
Brian Rabinovitz
@bdr1177
Oct 18 2017 03:13 UTC
Okay. I gotta get up early so I'm calling it a night. Thanks for fixing so quickly @scottleibrand !
DaisyHK
@laurathiessen
Oct 18 2017 03:14 UTC
Can't seem to upgrade to 0.6.0, is that related to the nogit issues?
Scott Leibrand
@scottleibrand
Oct 18 2017 03:14 UTC
sorry for breaking it, and thanks for testing the fix quickly
Jacob H
@jdhigh
Oct 18 2017 03:15 UTC
ok. Something happened. I did openaps first-upload and openaps upload-ns-status and now it's showing last data received 9 min ago.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:15 UTC
@laurathiessen give us 5 minutes to finish testing and merging this fix
assuming your problem is that oref0-pump-loop won't run
if you're getting a different error, please post it
or, if it's the same error as @bdr1177 got, I can go ahead and merge the fix based on his feedback and you can try again
LMK either way
Brian Bonner
@itsn0tm3
Oct 18 2017 03:18 UTC
@scottleibrand Still waiting for them both to start working, they're both still attempting to load up the files. However the edison I think may be stuck since it looks like this:
Starting pump-loop at Tue Oct 17 21:15:02 MDT 2017:
Waiting up to 4 minutes for new BG: date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.date: invalid date ‘@’
.
The pi currently looks like:
```
Starting pump-loop at Tue 17 Oct 21:18:11 MDT 2017:
Waiting up to 4 minutes for new BG: date: invalid date ‘@’
ls: cannot access monitor/pump_loop_completed: No such file or directory
Radio ok. Listening: .No pump comms detected from other rigs
Profile less than 60m old. date: invalid date ‘@’
temp_basal.json less than 5m old. date: invalid date ‘@’
pumphistory more than 5m old. : Refreshed pumphistory    at node.js:906:3
 and meal.json
Could not parse input data:  [SyntaxError: /home/pi/myopenaps/monitor/glucose.json: Unexpected end of input]
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
enact/enacted.json: cat: enact/enacted.json: No such file or directory
Error, retrying
pump_loop_completed more than 15m old; waiting for 30s silence before mmtuning
Radio ok. Listening: .
DaisyHK
@laurathiessen
Oct 18 2017 03:20 UTC
image.png
can't even get it to install 0.6.0
image.png
Brian Bonner
@itsn0tm3
Oct 18 2017 03:24 UTC
well the good news is the edison kept going, the bad news is they're both still complaining about the date now
Scott Leibrand
@scottleibrand
Oct 18 2017 03:26 UTC
you're getting CGM data from NS, or ?
@laurathiessen what error do you get when you try to install the latest dev?
Brian Bonner
@itsn0tm3
Oct 18 2017 03:27 UTC
yeah NS, just moved over to Heroku from Azure since I ran into that Azure bug which was why I re-ran the setup script and then hit your bug ha
so I went from one bug to the next.. oh well
Scott Leibrand
@scottleibrand
Oct 18 2017 03:27 UTC
ok, check your NS URL to start with, and then check ns-loop.log
I'm gonna go ahead and merge the fix
Brian Bonner
@itsn0tm3
Oct 18 2017 03:28 UTC
go for it, I think it's just me
hmm ns-loop looks ok to me
Starting oref0-ns-loop at Tue Oct 17 20:01:01 MDT 2017:
Found recent valid BG:
{"glucose":139,"dateString":"2017-10-17T19:56:19-06:00"}
Refreshed temptargets: {"target":null,"duration":null}
Refreshed carbhistory; COB: 64
No recent suggested.json found; last updated No recent treatments to upload
Completed oref0-ns-loop at Tue Oct 17 20:01:29 MDT 2017

Starting oref0-ns-loop at Tue Oct 17 20:02:02 MDT 2017:
Found recent valid BG:
{"glucose":139,"dateString":"2017-10-17T19:56:19-06:00"}
Refreshed temptargets: {"target":null,"duration":null}
Refreshed carbhistory; COB: IOB not found
Scott Leibrand
@scottleibrand
Oct 18 2017 03:30 UTC
that's old
from 30m ago
Brian Bonner
@itsn0tm3
Oct 18 2017 03:30 UTC
oh you're right
hmm the ns host and api secret are both correct
DaisyHK
@laurathiessen
Oct 18 2017 03:33 UTC
We try to install with the "deprecated" instructions and it ends up back at 0.5.2, when trying the new page the error is that it says already up to date, as in the picture.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:34 UTC
@laurathiessen can you show what you're actually running to update it? I just saw you checking what's installed
@itsn0tm3 is your ns-loop.log still updating with more recent runs?
DaisyHK
@laurathiessen
Oct 18 2017 03:35 UTC
image.png
Following this set of instructions here. Instead of updating, it stays on the old version. Doesn't seem to error though.
Brian Bonner
@itsn0tm3
Oct 18 2017 03:36 UTC
@scottleibrand no it doesn't seem to be updating at all
tryign to manually run the ns loop now
Scott Leibrand
@scottleibrand
Oct 18 2017 03:36 UTC
@laurathiessen can you show me the actual output of when you try to run the update and "doesn't seem to error"?
Brian Bonner
@itsn0tm3
Oct 18 2017 03:37 UTC
it just keeps doing this:
Radio ok. Listening: .No pump comms detected from other rigs
Profile less than 60m old. date: invalid date ‘@’
temp_basal.json more than 5m old.
Temp refreshed
Could not parse input data:  [SyntaxError: /root/myopenaps/monitor/glucose.json: Unexpected end of input]
determine-basal://text/shell/enact/suggested.json
reporting enact/suggested.json
enact/enacted.json: cat: enact/enacted.json: No such file or directory
Error, retrying
pump_loop_completed more than 15m old; waiting for 30s silence before mmtuning
Radio ok. Listening: ....................No pump comms detected from other rigs
Listening for 30s silence before mmtuning: .No pump comms detected from other rigs
Scott Leibrand
@scottleibrand
Oct 18 2017 03:37 UTC
@itsn0tm3 yeah, that's what you'll get until you get a glucose.json
so I would focus on ns-loop at this point, not pump-loop
Brian Bonner
@itsn0tm3
Oct 18 2017 03:37 UTC
how do you manually run the ns-loop?
Scott Leibrand
@scottleibrand
Oct 18 2017 03:38 UTC
oref0-ns-loop
Brian Bonner
@itsn0tm3
Oct 18 2017 03:38 UTC
pi@maria:~/myopenaps $ oref0-ns-loop
-bash: oref0-ns-loop: command not found
Scott Leibrand
@scottleibrand
Oct 18 2017 03:38 UTC
npm list -g oref0
Brian Bonner
@itsn0tm3
Oct 18 2017 03:39 UTC
root@megan:~/myopenaps# npm list -g oref0
/usr/local/lib
└── oref0@0.5.5
Scott Leibrand
@scottleibrand
Oct 18 2017 03:39 UTC
oh, this is an old rig
Brian Bonner
@itsn0tm3
Oct 18 2017 03:39 UTC
that's what both the pi and edison are reporting
DaisyHK
@laurathiessen
Oct 18 2017 03:39 UTC
image.png
Brian Bonner
@itsn0tm3
Oct 18 2017 03:40 UTC
the oref0-setup script should have updated that shouldn't it?
Scott Leibrand
@scottleibrand
Oct 18 2017 03:40 UTC
@laurathiessen ok, I would delete and re-clone your oref0 repo
cd ~/src/ && rm -rf oref0 && git clone https://github.com/openaps/oref0.git
and then re-do the other steps
@itsn0tm3 you're on master, so it'll just get the latest master
Brian Bonner
@itsn0tm3
Oct 18 2017 03:42 UTC
hmm actually I wasn't, I moved over to your test
so it should have updated to that
I just went ahead and moved it back over to dev
running the npm global install stuff now
Scott Leibrand
@scottleibrand
Oct 18 2017 03:42 UTC
not sure, but that sounds like the right action
Brian Bonner
@itsn0tm3
Oct 18 2017 03:43 UTC
yep doing those exact steps now
Scott Leibrand
@scottleibrand
Oct 18 2017 03:43 UTC
@laurathiessen you can also use http://openaps.readthedocs.io/en/latest/docs/Customize-Iterate/update-your-rig.html - the backup directory stuff isn't needed any longer
@laurathiessen either will work though, once we fix that npm error
DaisyHK
@laurathiessen
Oct 18 2017 03:44 UTC
Yeah, we had tried that too. But it also had issues. Having more problems trying to clear out
Jacob H
@jdhigh
Oct 18 2017 03:44 UTC
My NS info is still really slow to update. I tried this: Starting oref0-ns-loop at Tue Oct 17 23:38:59 EDT 2017:
No recent valid BG found. Most recent:
{"glucose":158,"dateString":"2017-10-17T23:25:00-04:00"}
Refreshed temptargets: {"target":80,"duration":60}
Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Warning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
Refreshed carbhistory; COB: 14
Could not require: /root/myopenaps/monitor/reservoir.json { [Error: Cannot find module '/root/myopenaps/monitor/reservoir.json'] code: 'MODULE_NOT_FOUND' }
{"BG":156,"IOB":2.816,"rate":2.1,"duration":30,"units":0.1}
No recent treatments to upload
Completed oref0-ns-loop at Tue Oct 17 23:39:30 EDT 2017
DaisyHK
@laurathiessen
Oct 18 2017 03:45 UTC
image.png
Scott Leibrand
@scottleibrand
Oct 18 2017 03:45 UTC
@laurathiessen looks like your rig's file system is corrupted. I'd re-flash it at this point.
DaisyHK
@laurathiessen
Oct 18 2017 03:46 UTC
okay, thanks. will do
Scott Leibrand
@scottleibrand
Oct 18 2017 03:47 UTC
@jdhigh MDT CGM? If so, that being slow to update is a know issue that no one has really tried to tackle yet. We only have Dexcom ourselves, so can't really help.
Jacob H
@jdhigh
Oct 18 2017 03:47 UTC
Yup. MDT.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:47 UTC
someone should probably create a github issue if there isn't one already and start coordinating everyone's efforts in one place
I haven't been keeping very close track
Jacob H
@jdhigh
Oct 18 2017 03:50 UTC
I wonder why the CGM type should have any effect on update speed? Openaps clearly knows my BG etc- it's putting the info into the openaps pill.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:53 UTC
because the MDT CGM setup is subtly broken and no one has taken the time to understand and fix it yet
Jacob H
@jdhigh
Oct 18 2017 03:53 UTC
lol
Scott Leibrand
@scottleibrand
Oct 18 2017 03:54 UTC
now everyone using MDT CGM please stand in a circle and point your finger in any direction you please. that person is the one who hasn't fixed it yet. ;-)
Brian Bonner
@itsn0tm3
Oct 18 2017 03:55 UTC
ha ouch
Jacob H
@jdhigh
Oct 18 2017 03:55 UTC
Aye! Well, at least I know it's not just me.
Scott Leibrand
@scottleibrand
Oct 18 2017 03:55 UTC
bystander effect in full force :-p
Brian Bonner
@itsn0tm3
Oct 18 2017 03:56 UTC

guiltyOfThat

alrighty all rigs up to date, re-ran the setup script, lets see how all 3 do now
Scott Leibrand
@scottleibrand
Oct 18 2017 03:56 UTC
:+1:
Jacob H
@jdhigh
Oct 18 2017 04:01 UTC
Ok well one final observation on the MDT CGM slowness. It seems that when I run openaps first-upload that NS updates.
Aaron Michelson
@Aaybob
Oct 18 2017 04:03 UTC
@jdhigh CGM data goes into the pump. Does the pump control the data? I haven't used Medtronic CGM for five, six years.
Jacob H
@jdhigh
Oct 18 2017 04:04 UTC
I'm not sure what you mean by "does the pump control"? The pump hosts the data, and I can look at the pump and see the BG and history of treatments, etc.
Aaron Michelson
@Aaybob
Oct 18 2017 04:04 UTC
You're way ahead of me.
I thought maybe the pump wasn't releasing the data...
Jacob H
@jdhigh
Oct 18 2017 04:06 UTC
AFAIK, the rig gets all the data up-to-the-minute.
And I can see all that data in the OpenAPS pill every other minute or so.
<<off to zzz
Brian Bonner
@itsn0tm3
Oct 18 2017 04:09 UTC
@scottleibrand Thanks for the help, both the pi and edison are running it so looks like the fix is working
Scott Leibrand
@scottleibrand
Oct 18 2017 04:29 UTC
@itsn0tm3 cool, thanks for confirming
Diadon81
@Diadon81
Oct 18 2017 04:50 UTC
I can confirm a problem with NS glucose data update
Scott Leibrand
@scottleibrand
Oct 18 2017 04:50 UTC
is this a new problem, or longstanding?
Diadon81
@Diadon81
Oct 18 2017 04:54 UTC
New one after my last update from dev
Scott Leibrand
@scottleibrand
Oct 18 2017 04:54 UTC
ok, then may be related to oref0-ns-loop.sh. what do you see?
Diadon81
@Diadon81
Oct 18 2017 04:57 UTC
There no any glucose data in NS only smb and boluses, openaps pill show 1m but NS pill show that last update was 7 hours ago. It was a moment when I update from dev
Scott Leibrand
@scottleibrand
Oct 18 2017 04:58 UTC
ok, I think I know the issue, one moment
Diadon81
@Diadon81
Oct 18 2017 04:58 UTC
Ok
Scott Leibrand
@scottleibrand
Oct 18 2017 05:02 UTC
just pushed a mdt-ns-loop branch
try checking that out and re-running oref0-setup
cd ~/src/oref0 && git fetch && git checkout mdt-ns-loop && git pull
Martin Haeberli
@mhaeberli
Oct 18 2017 05:44 UTC
running recent master. Edison goes offline when Edison goes into basement at college (think - office hours) - then when back in range of internet (per papertrail) can often take 1/2 hour to be fully online. Or T1 and openaps get separated (think visiting a study pal 3 doors down) - and 1/2 hour again. am i missing some magic?
Diadon81
@Diadon81
Oct 18 2017 05:58 UTC
Ok, I install from mdt-ns-loop
If I will get result before my plane fly, I will inform you about result
Anyway I will write something in next three hours))
Scott Leibrand
@scottleibrand
Oct 18 2017 06:12 UTC
@mhaeberli what do you mean by "be fully online"?
what do the logs show it doing during that half hour?
Diadon81
@Diadon81
Oct 18 2017 06:13 UTC
Ok, as a preliminary report, it fix the problem
Scott Leibrand
@scottleibrand
Oct 18 2017 06:13 UTC
:+1:
Paul Dickens
@thebookins
Oct 18 2017 08:17 UTC
Anyone have any tips for reducing iPhone battery drain while the rig is Bluetooth tethered? It's consuming 90% of total usage and the phone needs to be charged multiple times a day. Not really practical for our 11 year old.
Interestingly it's only emerged as a problem over the last week or so.
Velibor Maric
@vebaba
Oct 18 2017 09:10 UTC
You should write model and iOS version. Some 6Ss are eligible for battery replacement. Many on iOS 11.0.x report various problems etc.
Paul Dickens
@thebookins
Oct 18 2017 09:32 UTC
Good point @vebaba. It's an iPhone SE, just over a year old, running iOS 10.3.3.
Paul Dickens
@thebookins
Oct 18 2017 09:51 UTC
It's also been really sudden: the phone would last a whole day at school easily; now he needs to charge it before lunch. Coconut battery says the phone still holds over 90% of its design capacity, although who knows how reliable that is?
Velibor Maric
@vebaba
Oct 18 2017 10:21 UTC
Huh, SE and 10.3.3. aren't combination for sudden battery drain
ZandMarc
@ZandMarc
Oct 18 2017 10:38 UTC
my rig, running dev, got into a loop, with this error:
Oct 18 08:56:59 localhost kernel: [204142.165854] pci_pm_runtime_suspend(): serial_hsu_pci_runtime_suspend+0x0/0xf0 returns -16
Started the moment I walked out the door this morning and left my wifi. Does anybody know what can be the cause of this?
parenthetic
@parenthetic
Oct 18 2017 11:06 UTC
My OpenAPS pill in Nightscout has stopped working. I get "Device Undefined". There's still comms between rig, pump, and NS (temp basal rates are shown), but no details available in the pill (which I'd like to have back!). I've tried restarting the webapp and disabling the openaps plugin and then re-enabling it, but to no avail. I've checked the ns-loop.log and nothing seems amiss. If anyone has any pointers for what to check/try next I would very much appreciate it. Thanks!
Iain
@iainct
Oct 18 2017 11:14 UTC
This message was deleted
cameronrenwick
@cameronrenwick
Oct 18 2017 11:27 UTC
@parenthetic have you checked to see how much space you have in heroku? If you're looping fine yet get undefined, it's likely a database storage limit you've hit. The docs describe well how to first donate your data to openhumans and then how do selectively delete old data, freeing up space
parenthetic
@parenthetic
Oct 18 2017 12:56 UTC
@cameronrenwick I've got my NS site set up no Azure. I know it was recommended that Heroku was going to have more space/throughput available for a free app service and that the pricing was better than Azure if/when you went past the limits, but Azure seems to have upped their limits. Back in Feb or March I had my webapp shut down for a few hours once or twice, but then never bumped up against any limits. I haven't looked into donating my data to openhumans, but it's something I've been meaning to do. And I'll poke around my Azure webapp some more to see if I'm up against any limits, but I don't think I am. Thanks for your thoughts on this...
cameronrenwick
@cameronrenwick
Oct 18 2017 12:59 UTC
@parenthetic hey np.. I just know when I've been looping fine but had NS show undefined (for a prolonged period not just a blip) it's been due to storage. Glad to hear that Azure has bumped up their storage... didn't like getting those bills!
Jacob H
@jdhigh
Oct 18 2017 13:42 UTC
I have MDT CGM. I'm on the latest dev. I pulled the latest merge on the 16th (2 nights ago.) Since then, NS doesn't track my BG anymore. I feel that something from that merge broke it. If I do "openaps first-upload", then NS updates, but doesn't keep updating. Does anyone else have this problem?
Jacob H
@jdhigh
Oct 18 2017 13:50 UTC
^ Just noticed that earlier in the night @Diadon81 reported same problem. I'll try @scottleibrand 's fix. cd ~/src/oref0 && git fetch && git checkout mdt-ns-loop && git pull
parenthetic
@parenthetic
Oct 18 2017 14:17 UTC
@cameronrenwick I'm only using 14% of available storage on Azure. Thanks for your help, though!
ejoe132
@ejoe132
Oct 18 2017 14:28 UTC

trying to restart my rig and start from scratch, have the ediosn plugged in, it gets to "type root pasword or press control-d for maintenance, I type my password, then try to type reboot but it doesnt type it out then gives me this and just sits * PSH loader *
PCM page cache size = 192 KB
Cache Constraint = 0 Pages
Arming IPC driver ..
Adding page store pool ..
PagestoreAddr(IMR Start Address) = 0x04899000
pageStoreSize(IMR Size) = 0x00080000

Ready to receive application

Not sure maybe just time for new hardware

@ejoe132 That can be a symptom of inadequate power. Make sure you're using a beefy USB cable and a high-current USB port (like a wall charger, rather than a computer's USB port).
ejoe132
@ejoe132
Oct 18 2017 14:49 UTC
@ecc1 I have both USB in and the battery hooked up, so I think it should have plenty of power. not quite sure where to go with it as its pretty unresponsive
Try what I suggested, powering (only) from a wall charger with as thick a USB cable as you have, to eliminate some variables.
Velibor Maric
@vebaba
Oct 18 2017 15:22 UTC
Also, check whether Edison sits correctly (if power was an issue)
garykidd
@garykidd
Oct 18 2017 15:43 UTC
one thing I noticed about physical connection and how the edison may sit...is that the standoff's may need to be positioned such that hexagon shape is flat against the connector in order to provide enough room for the chip to sit deep enough in the connector...if the point of the hexagon shape face the connector--it could interfere with a nice solid connection.....
this though, of course, may not be your issue....
Diabetic_Dude
@diabetic_dude_twitter
Oct 18 2017 16:24 UTC
More power is always the answer
IMG_5415.jpg
Scott Leibrand
@scottleibrand
Oct 18 2017 16:32 UTC
@thebookins try turning off phone wifi briefly: your phone might be doing a wifi hotspot without you realizing it.
Diabetic_Dude
@diabetic_dude_twitter
Oct 18 2017 16:46 UTC
@thebookins what I do when I get battery drain issues is enable battery saver mode when I was testing the beta of IOS 11 i would have to charge 3+ times a day. I would go into settings, then battery, then enable the low power mode. That will disable all the background services that may be sucking up your battery.
(I would do that after a fresh charge )
is my board dead?? I keep getting this error
mmtune: KeyError: 'mmtune'
Usage: grep [OPTION]... PATTERN [FILE]...
Try 'grep --help' for more information.
Usage: grep [OPTION]... PATTERN [FILE]...
Try 'grep --help' for more information.
No wait required.
Scott Leibrand
@scottleibrand
Oct 18 2017 17:01 UTC
@diabetic_dude_twitter did you recently reinstall? that looks like it maybe didn't complete
Diabetic_Dude
@diabetic_dude_twitter
Oct 18 2017 17:02 UTC
was getting this error yesterday
do I killed my SRC folder renamed my openaps folder
and installed again
and started back up with the same error
wait.
you are right, I got that certificate error again because of the wrong date
I would bet that killed the install
Diabetic_Dude
@diabetic_dude_twitter
Oct 18 2017 17:07 UTC
the certificate error when trying to pull down an updated version of bluez
Scott Leibrand
@scottleibrand
Oct 18 2017 17:07 UTC
yeah, if it didn't ask you about cron at the end, it didn't finish
check your date and re-run
jaylagorio
@jaylagorio
Oct 18 2017 17:29 UTC
@scottleibrand Maybe a message at the end, "You've successfully completed the OpenAPS setup process!" ?
Scott Leibrand
@scottleibrand
Oct 18 2017 17:31 UTC
:+1: PR please!
Dana Lewis
@danamlewis
Oct 18 2017 18:11 UTC
(if anyone hadn't seen it elsewhere: https://twitter.com/danamlewis/status/920710209777016832)
jaylagorio
@jaylagorio
Oct 18 2017 18:29 UTC
@danamlewis Twitter and t.co are blocked where I am at the moment - mind posting what you're linking to in that tweet?
Dana Lewis
@danamlewis
Oct 18 2017 18:31 UTC
@jaylagorio np: https://diyps.org/2017/10/18/more-open-innovation-coming-soon/ (about the jdrf opening protocols announcement)
jaylagorio
@jaylagorio
Oct 18 2017 18:31 UTC
Thanks!
Dana Lewis
@danamlewis
Oct 18 2017 18:32 UTC
btw, for anyone who needs a reminder - dev is highly action packed right now. just merged the zero-temp prediction line stuff, which is now noted in the PR. It'll also require a NS update to see the line changes:
Screen Shot 2017-10-18 at 11.30.34 AM.png
(dev PR with all the notes: openaps/oref0#707)
Dana Lewis
@danamlewis
Oct 18 2017 18:40 UTC
also re: eliminating bolus snooze:
Screen Shot 2017-10-18 at 11.37.03 AM.png
big thanks to everyone who's been testing dev & PRs to dev & providing rapid feedback on what's working/not!
Marco
@CaptainBalou
Oct 18 2017 18:48 UTC

just pushed a mdt-ns-loop branch
try checking that out and re-running oref0-setup
cd ~/src/oref0 && git fetch && git checkout mdt-ns-loop && git pull

Had a problem with NS upload after dev „git pull“ stuff yesterday. I also tried the mdt-ns-loop branch. This fixes my problem. Is there any need to comment a ticket or something else?

Dana Lewis
@danamlewis
Oct 18 2017 18:49 UTC
@CaptainBalou it got merged last night so should be good to go for everyone in dev moving forward now - thanks for confirming it works! :+1:
Marco
@CaptainBalou
Oct 18 2017 18:51 UTC
Hi Dana. You merged last night? So means this change should be in dev already? I compared those two branches and saw this difference still. Shouldn’t it show up no differences after merging it with dev? confused
scottleibrand @scottleibrand double-checks
Scott Leibrand
@scottleibrand
Oct 18 2017 18:52 UTC
oops, too many fixes in one evening
Dana Lewis
@danamlewis
Oct 18 2017 18:52 UTC
heh good catch. @scottleibrand forgot to open the PR, but told me it was merged :) fixing now!
Marco
@CaptainBalou
Oct 18 2017 18:53 UTC
:-)
Scott Leibrand
@scottleibrand
Oct 18 2017 18:53 UTC
openaps/oref0#730
Dana Lewis
@danamlewis
Oct 18 2017 18:53 UTC
@CaptainBalou comment there and it can get merged :D
Brian Rabinovitz
@bdr1177
Oct 18 2017 18:53 UTC
@danamlewis I guess I'm still a little confused about what removing bolus snooze means for those of us who don't use SMB or UAM. I'm concerned that this will mean zero temping right after a meal, leading to unnecessary highs.
Marco
@CaptainBalou
Oct 18 2017 18:53 UTC
So do you want me to write a comment?
Ah ok. I’ll do.
Scott Leibrand
@scottleibrand
Oct 18 2017 18:54 UTC
@bdr1177 you shouldn't get zero temping as long as you enter carbs and it can accurately predict your bolus won't send you low
if you do, I want to fix that instead of continuing to bandaid it with bolus snooze

one other note for anyone upgrading dev. If you get:

error: Your local changes to the following files would be overwritten by merge:
 bin/oref0-upload-entries.sh
Please, commit your changes or stash them before you can merge.
Aborting

then just do what it suggests and run git stash and try again. that was from us forgetting to make that oref0-upload-entries.sh executable. npm helpfully does that for you, which is great until you want to pull an update. once you pull the new version of dev with the permissions set properly, you won't get that error again until the next time we add a file and forget to make it executable. :)

Brian Rabinovitz
@bdr1177
Oct 18 2017 18:56 UTC
I'll test it out and provide some feedback. I know I was having some problems with this a few days ago that I noted here and you said the latest dev may have fixed it. I'm on the latest dev now, so I'll look out for this issue.
Dana Lewis
@danamlewis
Oct 18 2017 18:57 UTC
thx @CaptainBalou :) now it's in dev :D
Marco
@CaptainBalou
Oct 18 2017 18:57 UTC
Oh that was a fast way. ;-)
So if everything doesn’t work well they see it was me who said it works. :-D
Dana Lewis
@danamlewis
Oct 18 2017 18:57 UTC
:)
Marco
@CaptainBalou
Oct 18 2017 18:58 UTC
Awesome. ;-)
Scott Leibrand
@scottleibrand
Oct 18 2017 18:58 UTC
@CaptainBalou you and @Diadon81 I believe :)
Dana Lewis
@danamlewis
Oct 18 2017 18:58 UTC
hehe yea. and so people know we're not just merging in stuff randomly just because. helpful to also then if we find another break know who can test which stuff!
Marco
@CaptainBalou
Oct 18 2017 18:58 UTC
Ok, so I go back to dev otherwise I get stuck in a deleted branch some day.
Dana Lewis
@danamlewis
Oct 18 2017 18:59 UTC
:+1: yep
Marco
@CaptainBalou
Oct 18 2017 19:00 UTC
Every time a problem because no one of you uses MDT, isn’t it? Changing code for things you do not have is hard. Is this an issue I could look into with bash skills or is other code involved?
Scott Leibrand
@scottleibrand
Oct 18 2017 19:00 UTC
bash skills are an excellent starting point for this
what we need here is for someone to take the stuff in the ns-loop openaps alias for MDT CGM and port it into the oref0-ns-loop bash script in a way that works with both CGM types
Marco
@CaptainBalou
Oct 18 2017 19:01 UTC
Starting point…and getting stuck after 30 minutes because I am heading java or whatever code? ;-)
Dana Lewis
@danamlewis
Oct 18 2017 19:01 UTC
@CaptainBalou (I'm the one with t1d, so correct, since I'm using Dexcom G4, @scottleibrand doesn't have a way to test on medtronic, so it's great for folks like you are helping test on other CGM types :))
Scott Leibrand
@scottleibrand
Oct 18 2017 19:01 UTC
so maybe some sort of extra argument on oref0-ns-loop for MDT CGM, and then add the extra MDT CGM upload stuff to the script conditional on that argument
if you know bash, hardest part will be reverse engineering what all the ns-loop alias (and the other openaps aliases and reports it calls) actually do
Marco
@CaptainBalou
Oct 18 2017 19:02 UTC
Yes Dana I love to help but skills are limited. bash is ok but nothing else. ;-)
Dana Lewis
@danamlewis
Oct 18 2017 19:03 UTC
that's more than I had when I started :)
Scott Leibrand
@scottleibrand
Oct 18 2017 19:03 UTC
no JS or python required for this one (although both are fairly easy to learn once you know some sort of coding)
Marco
@CaptainBalou
Oct 18 2017 19:03 UTC
Yes the „what does some aliases do“ and why is hard. I tried it last time while searching the preferences.json error I had. I was so confused...
Scott Leibrand
@scottleibrand
Oct 18 2017 19:04 UTC
openaps alias show foobar and openaps report show foobar.json --cli are your friends there
my goal with oref0-ns-loop and oref0-pump-loop is to get away from using openaps reports and aliases for anything other than pump comms, and change all other stuff into direct calls to the appropriate process from within the bash script
Marco
@CaptainBalou
Oct 18 2017 19:08 UTC
I try to read this chat lines a few times and try to see if it makes sense for me. ;-)
Scott Leibrand
@scottleibrand
Oct 18 2017 19:08 UTC
it won't at first, but once you poke around you can figure it out the same way we did at first. :)
Marco
@CaptainBalou
Oct 18 2017 19:09 UTC
Shoutout to all: If someone can fix this in some seconds please don’t wait for me to get it. :-D
regdubbed
@regdubbed
Oct 18 2017 19:29 UTC
So, we've been looping for 3 days now (Veo754, edison rig), and some questions came up:
Q1: When looping offline, there seems to be no bolus/carb data availiable to the loop. Is this "normal"?
Martin Haeberli
@mhaeberli
Oct 18 2017 19:31 UTC
@scottleibrand logs show it running loop, but getting errors pulling pump data (something like ‘NoneType’ - eventually, mmtune happens, then a little later it will be back in ‘enacted’ - so by ‘fully online’ I mean - actively controlling basals successfully. On a related note, sometimes will get stale BG, by 20 or 30 minutes, even when rig is reporting everything else and BGs in NightScout are current. From time to time, I can ‘fix’ those (WHEN I can remotely reach the rig), by doing a curl request for current bg from her Nightscout instance from within the rig. I’ll try to pull log excerpts and share ...
Dana Lewis
@danamlewis
Oct 18 2017 19:31 UTC
@regdubbed A1: it depends on how you're entering the data. if the rig doesn't have internet, it won't pull anything entered into NS. So for offline looping, you'll (for now) want to enter carbs and bolus into the pump directly.
that always works for getting data to the rig, off or online. when the rig has internet, you can choose to do the carb entry from a plethora of other ways.
regdubbed
@regdubbed
Oct 18 2017 19:32 UTC
We entered the data into the pump using the bolus wizzard. But looking at the logs, it seems the loop doesn't see it
Samuchco
@Samuchco
Oct 18 2017 19:34 UTC
I need to change pumps, from 522 to 722. Anyone knows an easy way to change just the serial number on the Oref0?
Dana Lewis
@danamlewis
Oct 18 2017 19:34 UTC
@regdubbed can you give an example of what you entered? Carbs need to go in with a 0.1 (or smallest increment for your pump) bolus
Scott Leibrand
@scottleibrand
Oct 18 2017 19:34 UTC
@mhaeberli openaps/oref0#728 might fix the stale BG when connectivity is flaky: I'd pull the latest dev
jaylagorio
@jaylagorio
Oct 18 2017 19:34 UTC
@Samuchco Just run setup again with the new serial number
Dana Lewis
@danamlewis
Oct 18 2017 19:34 UTC
@Samuchco nano your run-again file, change the SN, save the file, then bash runagain.
remember to re-update your preferences (edit-pref) after running setup
Samuchco
@Samuchco
Oct 18 2017 19:36 UTC
@danamlewis what is "bash runagain" ?
regdubbed
@regdubbed
Oct 18 2017 19:36 UTC
Just entered 18g /1u into the bolus wizard.
With the rig online that worked fine
Dana Lewis
@danamlewis
Oct 18 2017 19:37 UTC
where are you viewing your logs? are you talking about plugging into your rig and seeing the logs there? or saying it's not showing up in NS?
regdubbed
@regdubbed
Oct 18 2017 19:37 UTC
Is the bolus logged somewhere on the rig? I didn't find it
No NS wouldn't make sense since it's offline
Normally I see the IOB and COB values going up
Dana Lewis
@danamlewis
Oct 18 2017 19:40 UTC
can you paste your logs, since you said you just did the bolus wizard entry? let's take a look there
regdubbed
@regdubbed
Oct 18 2017 19:43 UTC
enacted.json: {"insulinReq":0,"bg":108,"reservoir":null,"temp":"absolute","snoozeBG":225,"carbsReq":3,"predBGs":{"IOB":[108,106,102,96,88,78,66,52,39,39,39,39
,39]},"minPredBG":999,"IOB":0.979,"reason":"COB: 0, Dev: -7, BGI: -0.87, ISF: 250, Target: 120, minPredBG -143, IOBpredBG 39; 3 add'l carbs req + 0m zero temp
; Eventual BG -144 < 110, bolus snooze: eventual BG range -144-225; setting current basal of 0.35 as temp. . Setting neutral temp basal of 0.35U/hr","COB":0,"
eventualBG":-144,"recieved":true,"duration":30,"tick":"+0","timestamp":"2017-10-15T15:46:52.311227","rate":0.35}
{"temp":"absolute","bg":98,"tick":-4,"eventualBG":-123,"snoozeBG":180,"insulinReq":0,"reservoir":null,"minPredBG":999,"predBGs":{"IOB":[98,94,87,78,67,54,39,3
9,39,39,39,39,39]},"COB":0,"IOB":0.938,"reason":"COB: 0, Dev: 13, BGI: -6.12, ISF: 250, Target: 120, minPredBG -130, IOBpredBG 39; 3 add'l carbs req + 0m zero
temp; Eventual BG -123 < 110, bolus snooze: eventual BG range -123-180, temp 0.35 ~ req 0.35U/hr. ","carbsReq":3}
Should'nt I see a COB here
Dana Lewis
@danamlewis
Oct 18 2017 19:46 UTC
yes. you said you're on a 754, right? what version of oref0 are you running?
Scott Leibrand
@scottleibrand
Oct 18 2017 19:47 UTC
The important part of the logs for this is at the top, right after starting pump-loop
regdubbed
@regdubbed
Oct 18 2017 19:47 UTC
top right?
Scott Leibrand
@scottleibrand
Oct 18 2017 19:48 UTC
At the top, right after
Dana Lewis
@danamlewis
Oct 18 2017 19:48 UTC
at the top, right after it says "starting pump-loop"
Diabetic_Dude
@diabetic_dude_twitter
Oct 18 2017 19:50 UTC
I found out what my problem is with me re running the oref0 setup and why it keeps failing. since I am connected VIA bluetooth and it wants to update bluetooth it kills my internet connection and tries to update bluez when I am on WiFi here at work the NTP server sucks and gives me the wrong date and also kills the connection to www.kernel.org because my date and time is wrong to fix this issue tried setting the wifi connection on my phone hotspot to priority 1 and did a reboot. still did not work. So manually tried to run a scan of my wifi networks. I saw the connection to my phone blip then go away, so then I tried to just get a new ip since I knew the phone was there and working... It seems that the Iphone update 11.2 changed my iphone's name from what I named it to Iphone I changed it back now all is running .
regdubbed
@regdubbed
Oct 18 2017 19:53 UTC

Starting pump-loop at Sun Oct 15 15:47:02 CEST 2017:
MDT CGM configured; not waiting
Radio ok. Listening: .No pump comms detected from other rigs

Attempting to retrieve MDT CGM data from pump
MDT CGM data retrieved
No new MDT CGM data to reformat

Profile less than 60m old. glucose.json newer than temp_basal.json. Temp refreshed
Temp duration 27m or more
glucose.json newer than pumphistory. : Refreshed pumphistoryWarning: clock input Invalid Date is unzoned; please pass clock-zoned.json instead
and meal.json
{"carbs":38,"boluses":4.6,"mealCOB":0,"currentDeviation":-1.33,"maxDeviation":1.08,"minDeviationSlope":-0.274}
{"iob":0.927,"activity":0.0012,"bolussnooze":0.973,"basaliob":-0.065,"netbasalinsulin":-0.05,"hightempinsulin":0.05,"microBolusInsulin":0,"microBolusIOB":0,"time":"2017-10-15T13:48:00.000Z","lastBolusTime":1508074945000}
{"delta":0,"glucose":108,"short_avgdelta":-2.02,"long_avgdelta":-1.12}
{"duration":29,"rate":0.35,"temp":"absolute"}
Basal unchanged: 0.35; target_bg unchanged: 120; sens unchanged: 250 (autosens ratio 1)
Carb Impact: -0.5 mg/dL per 5m; CI Duration: 0 hours; remaining 4h+ CI: 0 mg/dL per 5m
Accel. Carb Impact: 10 mg/dL per 5m; ACI Duration: 0 hours
UAM Impact: -0.5 mg/dL per 5m; UAM Duration: 0 hours
minPredBG: -127 minIOBPredBG: 39 avgPredBG: -127 COB: 0 carbs: 38
BG projected to remain above 110 for 0 minutes
bgUndershoot: 81 zeroTempDuration: 30 zeroTempEffect: 44 carbsReq: 3

Scott Leibrand
@scottleibrand
Oct 18 2017 19:55 UTC
So it is seeing 38g of carbs. What carbs have you entered over the last 6h or so?
Or feel free to PM me your NS URL if that's easier.
regdubbed
@regdubbed
Oct 18 2017 19:57 UTC
those 38 carbs have been there since 11AM
Paul Dickens
@thebookins
Oct 18 2017 19:57 UTC
Thanks for the battery drain suggestions. Is it likely to depend at all on cellular signal strength? We changed providers and sim the other day - the only thing we can think of that has changed.
Scott Leibrand
@scottleibrand
Oct 18 2017 20:02 UTC
If signal is weak it'll try really hard and burn lots of battery, ya.
@regdubbed this one might be hard to troubleshoot retrospectively. Is it reproducible? Could you take your rig offline with a console connection at some point to help figure it out?
Martin Haeberli
@mhaeberli
Oct 18 2017 20:09 UTC
@scottleibrand thx
regdubbed
@regdubbed
Oct 18 2017 20:10 UTC
Well... at least I could try. What would I want to look at then?
Scott Leibrand
@scottleibrand
Oct 18 2017 20:11 UTC
Whether the carbs show up in pumphistory.json, and then in meal.json.
And if necessary, running the oref0-meal command manually
It might be only getting run on NS carbhistory, not pumphistory. That works fine normally, but that might be one of the ways the MDT CGM stuff is subtly broken.
regdubbed
@regdubbed
Oct 18 2017 20:14 UTC
it's still in the pump:
{
"_type": "BolusWizard",
"bg": 11.0,
"bg_target_high": 13.0,
"_description": "BolusWizard 2017-10-15T15:42:24 head[2], body[15] op[0x5b]",
"timestamp": "2017-10-15T15:42:24",
"_body": "125000aafa6e000028000000002882",
"_head": "5b6e",
"unabsorbed_insulin_total": 0.0,
"correction_estimate": 0.0,
"_date": "98aa0f0f11",
"bolus_estimate": 1.0,
"bg_target_low": 11.0,
"carb_ratio": 17.0,
"food_estimate": 1.0,
"carb_input": 18,
"sensitivity": 25.0
},
but not in pumphistory.jso
Scott Leibrand
@scottleibrand
Oct 18 2017 20:15 UTC
Yeah. My question is why meal.json didn't have it at the time.
Sorry, it's missing in pumphistory.json? Or it just doesn't go that far back?
regdubbed
@regdubbed
Oct 18 2017 20:16 UTC
Doesn't go that far back
Scott Leibrand
@scottleibrand
Oct 18 2017 20:16 UTC
K
regdubbed
@regdubbed
Oct 18 2017 20:16 UTC
Ok, I 'll try to reproduce it
Jacob H
@jdhigh
Oct 18 2017 20:20 UTC
I'm excited to try the wifi-hotspot merge. openaps/oref0#716 Question: What app do you recommend on an android phone to ssh in?
Samuchco
@Samuchco
Oct 18 2017 20:22 UTC
Changing pump to new SN. I re-run the set up. Changed SN. Bash, and I'm getting this error message
Using /home/.rootfs/usr/local/lib/python2.7/dist-packages
Finished processing dependencies for decocare==0.1.0-dev
/root/src/oref0/ already exists; pulling latest
Updating fb20124..bcd546c
error: Your local changes to the following files would be overwritten by merge:
bin/oref0-upload-entries.sh
Please, commit your changes or stash them before you can merge.
Aborting
Couldn't pull latest oref0
Dana Lewis
@danamlewis
Oct 18 2017 20:23 UTC
@jdhigh have you been using xdripAPS already? there's already some good options for android users, but if you end up testing the android version of offline, it'll be helpful to have you help document the differences between each type of offline setup
@Samuchco do git stash then re-run
Samuchco
@Samuchco
Oct 18 2017 20:24 UTC
@danamlewis root@SammysAPS:~/myopenaps# git stash
fatal: Not a git repository (or any of the parent directories): .git
probablyneed to change dir
regdubbed
@regdubbed
Oct 18 2017 20:26 UTC
hmm, concerning pumphistory.json: shouldn't this have the last 100 entries?
Jacob H
@jdhigh
Oct 18 2017 20:26 UTC
@danamlewis I haven't. I don't know anything about xdrip; I have an MDT CGM and I just go to the NS site on my phone.
Samuchco
@Samuchco
Oct 18 2017 20:30 UTC
@danamlewis should I start fresh?
Scott Leibrand
@scottleibrand
Oct 18 2017 20:35 UTC
@Samuchco you're just in the wrong directory
cd ~/src/oref0 first
Samuchco
@Samuchco
Oct 18 2017 20:35 UTC
I thought so
thanks
Scott Leibrand
@scottleibrand
Oct 18 2017 20:36 UTC
@regdubbed pumphistory.json will be short. there is a pumphistory-24h.json, and then the two get merged into pumphistory-merged.json
Samuchco
@Samuchco
Oct 18 2017 20:44 UTC
@scottleibrand Do you want my son's data. He is on 0.6.0 Dev2 autotune and autosense. If yes, how?
Paul Dickens
@thebookins
Oct 18 2017 20:55 UTC
@scottleibrand it must be the low signal strength that's burning through battery on Bluetooth tether. Which makes battery life highly dependent on where you are. I wonder would it make sense for a future feature to connect to the bt pan only once every five minutes ( driven perhaps by pump-loop) and then drop the tether? Or do we need to interrogate NS more frequently than that? We are planning for our son being away on school camp with variable signal strength and nervous about the phone going flat while out and about.
mmadeva
@mmadeva
Oct 18 2017 21:33 UTC
Dear All! I started OpenAps three days ago with Intel Edison, G4 +receiver , I get BG data from Nightscout. My problem is that there is no visualization on Nightscout of carbs and bolus when I use Bolus Wizard. All other events like correction bolus, suspen/resume pump , settings changes are uploaded immediately on Nightscout. More over, all these events are listed like future treatment in Admin Tools page although they are with current date and time. I am concerned if carbs and bolus insulin are recorded in appropriate files for correct openaps calculations. I checked the date of all devices inc. of the rig-it is one and the same. Does anyone have the same experience?
Scott Leibrand
@scottleibrand
Oct 18 2017 21:52 UTC
@thebookins if you wanted to you could put a 3-4 minute sleep into oref0-ns-loop (or run the cron every 5m). it just wouldn't pick up BGs, carbs, and temptargets quite as fast
Paul Dickens
@thebookins
Oct 18 2017 21:54 UTC
Cool, I'll try that @scottleibrand , just want to make battery life a bit more deterministic.
Marco
@CaptainBalou
Oct 18 2017 22:16 UTC
@scottleibrand @danamlewis One script an alias for another bunch of scripts, openaps commands I am not aware of what the do, files I see getting not filled but I don’t know why - I don’t see which files are needed and how especially they are getting created. I am sorry but I cannot involve my bash skills without having someone who helps me getting in touch with all that. Sorry for that. It’s past midnight. cu guys.
Scott Leibrand
@scottleibrand
Oct 18 2017 22:28 UTC
ok, hit us up another time and we can walk you through how to disentangle all the openaps commands
PopcornPam
@PopcornPam_twitter
Oct 18 2017 23:11 UTC
Question about bluetooth tethering... I have noticed a pretty consistant pattern across rigs and phones that the phone can't connect to the rig unless I restart it. is it just me? Any tricks that have worked for people?
philipgo
@philipgo
Oct 18 2017 23:28 UTC
@PopcornPam_twitter Are you using an app like "Blue Car Tethering" to always allow bluetooth tethering?
PopcornPam
@PopcornPam_twitter
Oct 18 2017 23:29 UTC
no... never heard of it
should I be?
I don't want to tether at home or work where I have wifi
KS615
@KS615
Oct 18 2017 23:30 UTC
@PopcornPam_twitter It seems like phones vary widely. What phone are you using?
philipgo
@philipgo
Oct 18 2017 23:31 UTC
@PopcornPam_twitter Most Android phones disable bluetooth internet connection sharing after a certain time. This would explain why it can connect after the phone has been restarted. We use "Blue Car Tethering", it solved this problem for us. The app is free.
Scott Leibrand
@scottleibrand
Oct 18 2017 23:36 UTC
I think @PopcornPam_twitter is on iPhone
philipgo
@philipgo
Oct 18 2017 23:37 UTC
Oh ok, we had no consistent BT tethering results with iPhones unfortunately