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

13th
Sep 2018
Sacha M
@coolestkidsever
Sep 13 2018 00:48

Hi all, I've purchased a second hand / flashed edison rig, and am running through a new setup, but keep coming up to this error:


Restarting ntp (via systemctl): ntp.service.
bound to 192.168.2.6 -- renewal in 1623 seconds.

Wifi SSID: icehouse
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:--  0:00:19 --:--:--     0curl: (6) Could not resolve host: raw.githubusercontent.com

following directions from here: https://openaps.readthedocs.io/en/latest/docs/Build%20Your%20Rig/edison-install.html
Does anyone know how to resolve please?

Sacha M
@coolestkidsever
Sep 13 2018 01:20
rebooted a few times and got it working
phoenixflame611
@phoenixflame611
Sep 13 2018 03:32
image.png
Hi all, I have infrequently getting the following error that causes a 20-30 min drop in looping at its worst. any ideas on what this could be?
philipgo
@philipgo
Sep 13 2018 03:46
@alimhassam Does jubilinux 0.4.0 work with OpenAPS without any changes?
Scott Leibrand
@scottleibrand
Sep 13 2018 07:50
@phoenixflame611 which version of oref0? Need to see what your lines 567 and 568 are.
boc-the-git
@boc-the-git
Sep 13 2018 08:57

Hey folks, just been talking with Dana on Facebook about an issue I'm having. I'm running Pi+Explorer Hat with a 522 and mdt cgm, and oref0-version 0.7.0-dev [dev]
My issue is that I'm not getting any BG data uploaded to NS - absolutely no records getting loaded into 'entries'. I also have a potentially related issue, nightscout/cgm-remote-monitor#3904
In terms of getting the BGs uploaded, Dana advised running openaps first-upload, which gives me this:

root@Liverbird:~/myopenaps# openaps first-upload
[Errno 2] No such file or directory
openaps: 'first-upload' is not an openaps command. See 'openaps --help'.

Dana Lewis
@danamlewis
Sep 13 2018 09:05
That doesn't work on Dev, so tagging @juehv or @cluckj to see what the latest recommendation for fix or branch is for mdt CGM on dev to resolve.
Thx for hopping over here @BOC14!
boc-the-git
@boc-the-git
Sep 13 2018 09:23
No worries at all! As I say, potentially related that other issue as well, depending on whether that's a nightscout or openaps issue - my gut feel initially was NS but could be wrong.
alimhassam
@alimhassam
Sep 13 2018 12:29
@philipgo it works and it's nicer because it has one larger 3gb partition instead of 2 smaller ones. But it's a bit morew difficult to install as you need to press a key to stop at the uboot loader and enter a command.
Jens Heuschkel
@juehv
Sep 13 2018 13:18
@BOC14 the upload to nighscout is integrated in cgmupdate. You may check /var/log/openaps/cgm-loop.log and check if there is some upload error
I have to admit, I never thought about this initial upload. Could you try to install the master branch, call the initial upload and switch back to dev and see if it works ?
Dana Lewis
@danamlewis
Sep 13 2018 13:26
@juehv it will happen for every first time looper w/ Medtronic CGM, so we'll need to hopefully automate it with Go after we figure out how to do it. Thx for jumping in here!
Jens Heuschkel
@juehv
Sep 13 2018 13:44
@danamlewis sure. But first we need to find out if that solve the problem or if there is a different issue ;)
Raymond Richmond
@PedanticAvenger
Sep 13 2018 14:27
Looks like jubilinux.org is coming back to life. Slowly so I suspect they had a disaster recovery scenario. Timely reminder to have alternatives available. ;)
David
@davidkeddydb
Sep 13 2018 15:20
@danamlewis Once the comms stuff gets finalized over at OpenOmni, which I realize is getting closer and closer (excited for my T1d) are there plans for OpenAPS working with Omnipod?
Eric
@ecc1
Sep 13 2018 15:59

@davidkeddydb The good news is that the same CC111x radios on Explorer boards and TI sticks should work with Omni, with the new firmware that @ps2 has developed for RileyLink. The bad news is that someone will have to write the "driver" side, to run on the Edison or Pi and talk to it according to the new protocol.

It should be possible to extend my cc111x driver to do that, but it won't be me (although I'll be happy to help whoever volunteers)

Dana Lewis
@danamlewis
Sep 13 2018 16:44
^ that. there is also ongoing work to make Omnipod driver work for Android, so those interested can use AndroidAPS (with OpenAPS algo)
I haven't heard anyone start talking about work on a driver on Edison/Pi, so AAPS will probably be the earliest available thing using OpenAPS algo
n0rbertg
@n0rbertg
Sep 13 2018 17:11

^ brownie points for that link :smile: as a Pr

done

Dana Lewis
@danamlewis
Sep 13 2018 17:27
­čÄé­čÄŐ­čÄë
philipgo
@philipgo
Sep 13 2018 17:48
@alimhassam What command is it you have to enter while installing jubilinux 0.4.0
alimhassam
@alimhassam
Sep 13 2018 18:24
I wrote notes on the jubilinux gutter channel on what I had to do to make it work. All steps might not be required
alimhassam
@alimhassam
Sep 13 2018 18:31
Let me know if it works for you and maybe we can come up with easier simpler instructions. I don't fully understand why all the steps are needed yet, but I really like having just one partition so I don't mind v doing the extra step while flashing
jalonergan
@jalonergan
Sep 13 2018 20:31
image.png
Hello need help with updating maxsafebasal. I can't figure out how to raise the maxsafebasal to 0.75. I've read the docs, made adjustments accordingly, and re-ran setup. And it still had no impact to maxsafebasal. What else is limiting it? Autosen or Autotune?
Sacha M
@coolestkidsever
Sep 13 2018 21:06
@jalonergan do you have a max basal set in the pump?
jalonergan
@jalonergan
Sep 13 2018 21:07
@coolestkidsever yes, it is set at 2.0
jalonergan
@jalonergan
Sep 13 2018 21:14
@zgohr both multiplier adjusted. I'm going turn off autotune and see if that makes a difference.
boc-the-git
@boc-the-git
Sep 13 2018 21:31
@juehv No apparent issues in cgm-loop.log, other than the occasional timeout. It reads to me as if everything went well but it decided not to upload anything..
Starting oref0-mdt-update at Fri 14 Sep 07:18:35 AEST 2018:
Most recent cgm record: {"sgv":158,"dateString":"2018-09-14
07:08:00+10:00"}
updating cgm data
Listening for 1s: .2018/09/14 07:18:37 connected to CC111x
adio on /dev/spidev0.0
2018/09/14 07:18:37 setting frequency to 916.650
2018/09/14 07:18:37 listening for 1s
2018/09/14 07:18:38 timeout
No interfering pump comms detected from other rigs (this is
a good thing!)
Continuing oref0-pump-loop at Fri 14 Sep 07:18:38 AEST 2018
2018/09/14 07:18:39 PAPERTRAIL environment variable is not
et
2018/09/14 07:18:39 read 161 entries from monitor/cgm-mm-gl
cosedirty.json
2018/09/14 07:18:40 connected to CC111x radio on /dev/spide
0.0
2018/09/14 07:18:40 setting frequency to 916.650
2018/09/14 07:18:40 model 522 pump
2018/09/14 07:18:40 CGM clock difference = 15.663860798s
2018/09/14 07:18:40 retrieving records since 2018-09-14 07:
8:00
2018/09/14 07:18:43 history page 6: received fragment 6 aft
r 1 NAK
2018/09/14 07:18:45 writing CGM timestamp for page 6 and re
canning
2018/09/14 07:18:50 stopping CGM history scan at 2018-09-14
07:08:00
2018/09/14 07:18:50 6 CGM records
2018/09/14 07:18:50 0 Nightscout entries
2018/09/14 07:18:50 merging 161 old and 0 new entries
2018/09/14 07:18:50 161 merged entries from 2018-09-14 07:0
:00 to 2018-09-13 17:49:00
2018/09/14 07:18:50 161 trimmed entries from 2018-09-14 07:
8:00 to 2018-09-13 17:49:00
2018/09/14 07:18:50 wrote 161 entries to monitor/cgm-mm-glu
osedirty.json
2018/09/14 07:18:50 retrieving Nightscout records from last
1h0m25.303942932s
2018/09/14 07:18:57 looking for gaps in 2 Nightscout record
2018/09/14 07:18:57 uploading 0 entries to Nightscout
MDT CGM data retrieved
MDT New cgm data reformatted
Most recent cgm record: {"sgv":158,"dateString":"2018-09-14
07:08:00+10:00"}
Completed oref0-mdt-update at Fri 14 Sep 07:19:00 AEST 2018
Sorry for the spam above.. @juehv you mentioned trying to install master.. that's not something I'll get to for at least 12 but possibly 24/36 hours.. can I just confirm, you're suggesting to install master, run openaps first-upload, then switch back to dev?
Oh also note, it's legitimate that there is no new reading in the 10 minute period there, I'm currently starting a new sensor
Further back in the logs I can see things like 'uploading 1 entries to Nightscout' with no error messages following it..
Eric
@ecc1
Sep 13 2018 21:39
@BOC14 once there are new readings, try running the cgmupdate command with the -v option to see what gaps it thinks there are in NS and what it thinks it can upload. Right now it's saying none of the 161 (old) entries help to fill any of the gaps it's found in NS
Tore Bj├Şrndalen
@BjoTore_twitter
Sep 13 2018 21:49
@BOC14 check your ns-loop.log. If you find "Load high at Thu 13 Sep 23:24:43 CEST 2018: waiting up to 5m to continue". I see this a lot.
It will delay (for how long?) the NS upload while load is higher than 4
@ecc1 I'm also seeing this too often:
Sep 13 21:45:27 ToresRpiZero pump-loop.log: Sending ESC ESC, ESC ESC ESC ESC to exit any open menus before SMBing Couldn't mdt -f internal button esc esc - continuing
Sep 13 22:45:16 ToresRpiZero pump-loop.log: Sending ESC ESC, ESC ESC ESC ESC to exit any open menus before SMBing Couldn't mdt -f internal button esc esc - continuing
Sep 13 22:48:12 ToresRpiZero pump-loop.log: Sending ESC ESC, ESC ESC ESC ESC to exit any open menus before SMBing Couldn't mdt -f internal button esc esc - continuing
Sep 13 23:18:01 ToresRpiZero pump-loop.log: Sending ESC ESC, ESC ESC ESC ESC to exit any open menus before SMBing and bolusing 0.3 units
Sep 13 23:24:00 ToresRpiZero pump-loop.log: Sending ESC ESC, ESC ESC ESC ESC to exit any open menus before SMBing Couldn't mdt -f internal button esc esc - continuing
Sep 13 23:41:24 ToresRpiZero pump-loop.log: Sending ESC ESC, ESC ESC ESC ESC to exit any open menus before SMBing and bolusing 0.4 units
Tore Bj├Şrndalen
@BjoTore_twitter
Sep 13 2018 21:54
I'm using your new binaries btw
alimhassam
@alimhassam
Sep 13 2018 22:00
I wanted to try new binaries but didn't have time yesterday. Seems like something didn't get installed correctly for you @BjoTore_twitter
Can you try see if you have the mdt command in your path?
which mdt
Or just mdt
Tore Bj├Şrndalen
@BjoTore_twitter
Sep 13 2018 22:07
I was just going over that now, to see if I did something wrong during copying. So far it seems ok
root@ToresRpiZero:~/myopenaps# which mdt
/usr/local/bin/mdt
It works most of the time
Tore Bj├Şrndalen
@BjoTore_twitter
Sep 13 2018 22:20
I ran a test a little earlier:
root@ToresRpiZero:~/myopenaps# service cron stop
root@ToresRpiZero:~/myopenaps# kill 553^C
root@ToresRpiZero:~/myopenaps# killall -g oref0-pump-loop
root@ToresRpiZero:~/myopenaps# killall -g oref0-pump-loop
oref0-pump-loop: no process found
root@ToresRpiZero:~/myopenaps# mdt -f internal button esc esc
2018/09/13 22:51:58 connected to CC111x radio on /dev/spidev0.0
2018/09/13 22:51:58 setting frequency to 868.500
2018/09/13 22:52:00 waking pump
2018/09/13 22:52:12 no response to wakeup
root@ToresRpiZero:~/myopenaps# mdt -f internal button esc esc
2018/09/13 22:52:17 connected to CC111x radio on /dev/spidev0.0
2018/09/13 22:52:17 setting frequency to 868.500
2018/09/13 22:52:19 waking pump
2018/09/13 22:52:29 pressing EscButton
2018/09/13 22:52:29 pressing EscButton
2018/09/13 22:52:29 disconnecting CC111x radio on /dev/spidev0.0
root@ToresRpiZero:~/myopenaps# mdt -f internal button esc esc
2018/09/13 22:52:37 connected to CC111x radio on /dev/spidev0.0
2018/09/13 22:52:37 setting frequency to 868.500
2018/09/13 22:52:37 model 515 pump
2018/09/13 22:52:37 pressing EscButton
2018/09/13 22:52:37 pressing EscButton
2018/09/13 22:52:37 disconnecting CC111x radio on /dev/spidev0.0
Tore Bj├Şrndalen
@BjoTore_twitter
Sep 13 2018 22:28
I went over all the binaries and checked them with md5sum
Checking the tar ball.....
boc-the-git
@boc-the-git
Sep 13 2018 23:01
@ecc1 @BjoTore_twitter when running cgmupdate -v I'm getting "MEDTRONIC_PUMP_ID environment variable
s not set"
It looks like that error has some other discussions in this channel, so I'll give them a read and look to resolve that as my next step.
No sign of that error in the ns-loop either @BjoTore_twitter , though there was various mentions of invalid json. Should have more time in a few hours to spend digging deeper!
alimhassam
@alimhassam
Sep 13 2018 23:15
@BjoTore_twitter could your issue be that your frequency isn't mmtuned properly?
@BOC14 try running these 2 commands before running cgmupdate
cd ~/myopenaps
export MEDTRONIC_PUMP_ID=`grep serial pump.ini | tr -cd 0-9`
export MEDTRONIC_FREQUENCY=`cat monitor/medtronic_frequency.ini`
boc-the-git
@boc-the-git
Sep 13 2018 23:48

Thanks @alimhassam - after that, I get the following from cgmupdate -v:
root@Liverbird:~/myopenaps# cgmupdate -v
2018/09/14 09:38:39 PAPERTRAIL environment variable is not set
2018/09/14 09:38:40 connected to CC111x radio on /dev/spidev0.0
2018/09/14 09:38:40 setting frequency to 916.600
2018/09/14 09:38:40 model 522 pump
2018/09/14 09:38:41 CGM clock difference = 15.099470751s
2018/09/14 09:38:41 retrieving records since 2018-09-14 09:18:26
2018/09/14 09:38:42 glucosePage command required 1 retry
2018/09/14 09:38:46 stopping CGM history scan at 2018-09-14 09:18:00
2018/09/14 09:38:46 4 CGM records
2018/09/14 09:38:46 3 Nightscout entries from 2018-09-14 09:33:00 to 2018-09-14 09:23:00
[
{
"type": "sgv",
"date": 1536881580000,
"dateString": "2018-09-14T09:33:00+10:00",
"device": "openaps://Liverbird",
"sgv": 178,
"direction": "Flat"
},
{
"type": "sgv",
"date": 1536881280000,
"dateString": "2018-09-14T09:28:00+10:00",
"device": "openaps://Liverbird",
"sgv": 178,
"direction": "Flat"
},
{
"type": "sgv",
"date": 1536880980000,
"dateString": "2018-09-14T09:23:00+10:00",
"device": "openaps://Liverbird",
"sgv": 180
}
]

This is my subsequent ns-looplog:
Starting oref0-ns-loop at Fri 14 Sep 09:41:04 AEST 2018:
No recent valid BG found. Most recent:
{"glucose":null,"dateString":null}
{"sgv":null,"dateString":null}
Latest NS temptargets: (NOT VALID JSON: empty)
Merging local temptargets: (NOT VALID JSON: empty)
Temptargets merged: (NOT VALID JSON: empty)
Refreshed carbhistory; COB: 0
{"batteryVoltage":4000,"battery":87}
No recent suggested.json found; last updated Sep 14 09:39
No new treatments to upload
Completed oref0-ns-loop at Fri 14 Sep 09:47:57 AEST 2018

Note there is still nothing in the entries table of my NS. I'm guessing that 'No recent valid BG found' may be relevant?

Tore Bj├Şrndalen
@ToreBj
Sep 13 2018 23:58
@alimhassam this is something I monitor frequently. Sometimes mmtune fails, but it keeps the previous result.