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

2nd
May 2017
garykidd
@garykidd
May 02 2017 00:05
@applehat what app are you referring to that displays NS info on android wear? I have iPhone and use android wear app on iPhone to interface with watch...
BDavis1960
@BDavis1960
May 02 2017 00:05
Network log 5:1.png
Had a few successful first days of looping with wifi, and yesterday tried BT tethering! However, when I returned home last night, my rig didn't connect. Can't seem to make that happen. Here are some logs. Any help is appreciated!
Loop log 5:1.png
Lin
@denglfddd
May 02 2017 00:48
root@localhost:~# ifup wlan0
ifup: interface wlan0 already configured
My rig looks and does not connect to WIFI?
cameronrenwick
@cameronrenwick
May 02 2017 00:55
anyone considered using a lipo like this one? https://elmwoodelectronics.ca/products/12895 I think it may be a consideration for my updated build. To be able to have several of them in a charger for plugging into a rig. Have to make certain of polarity but... just wondering if any of the 3D case pros here have thought about this. I imagine a rig that has one plugged in and a second slot as a spare... thoughts?
rsilvers129
@rsilvers129
May 02 2017 01:04
@cameronrenwick Yes, people use those. I don’t love the thought of just killing power to a unix computer. But it may be ok.
cameronrenwick
@cameronrenwick
May 02 2017 01:08
just thought I'd share a little find regarding wifi tethering on my Samsung Galaxy S7. I'm in Canada and it seems my carrier (Rogers) does not support bluetooth "tethering" (different from say connecting you phone via bluetooth to your car for hands free talking). To set up your hotspot go into settings/connections/mobile hotspot and tethering. Switch mobile hotspot "on". In your rig's wifi set-up script (vi /etc/wpa_supplicant/wpa_supplicant.conf) make sure the hotspot and password are included. When you go out of your home/office wifi environment, make sure you turn the hotspot to "on". Despite the fact that the Galaxy S7 asks for a bluetooth-like address (AA:BB:CC...) leaving this blank is fine and your rig will connect fine. I share this as bluetooth IS available on US S7 phones but not Canadian ones it seems.
@rsilvers129 I agree. Here's my thought tho, I've got my rig on a 2000 mah w a jst plug. Coming home it died (12 hr day) and I had a period of time where I was not looping. Had I had another battery I could plug in, I'd have done so. The jst connectors are pretty light duty and I don't really want to get into the habit of swapping those types of batteries. Possibly I could get the solder tab style round (AA-like) lipos I listed above and have some more robust connector on it to afford easier/safer swapping. I just think for me, I'm gonna need more mah's than 2000 for a full day's work. Maybe a 2500mah would be better but ..
rsilvers129
@rsilvers129
May 02 2017 01:13
Maybe just build a rig with 5000 mAh. I am building a 2000 and a 2500, but the 2000 is just if I go for a run or something.
cameronrenwick
@cameronrenwick
May 02 2017 01:15
so you've got 3 independent rigs with their own lipos depending on activity then?
rsilvers129
@rsilvers129
May 02 2017 01:18
I am making two. I plan to use the 2500 most of the time. I think I need two just to service/upgrade one while the other is working anyway. Also it seems like the parts are often on backorder, so it makes sense to have two sets.
You can also charge it while you are driving.
tynbendad
@tynbendad
May 02 2017 01:43
@cameronrenwick @danimaniac has an 18650 edison rig case https://github.com/danimaniac/OpenAPS-Explorer-Board-Edison-vented-case
rsilvers129
@rsilvers129
May 02 2017 01:47
Here is a new case I just finished for the 2000. http://www.thingiverse.com/thing:2288682
tynbendad
@tynbendad
May 02 2017 01:50
@rsilvers129 looks great, love the open window. still monitoring the 2500 battery life to see if we can maybe use a 2000 instead...
dougtakeuti
@dougtakeuti
May 02 2017 02:51
Hi, I just tried to update my RPI rig. updating oref0 went ok, but this is what I get for npm run global-install
pi@raspberrypi:~/src/oref0 $ npm run global-install

> oref0@0.5.0-dev global-install /home/pi/src/oref0
> npm install && sudo npm install -g && sudo npm link && sudo npm link oref0

npm ERR! Error: EPERM, chmod '/home/pi/src/oref0/node_modules/oref0/bin/bt-pan'
npm ERR!  { [Error: EPERM, chmod '/home/pi/src/oref0/node_modules/oref0/bin/bt-pan']
npm ERR!   errno: 50,
npm ERR!   code: 'EPERM',
npm ERR!   path: '/home/pi/src/oref0/node_modules/oref0/bin/bt-pan' }
npm ERR!
npm ERR! Please try running this command again as root/Administrator.

npm ERR! System Linux 4.1.13-v7+
npm ERR! command "/usr/bin/nodejs" "/usr/bin/npm" "install"
npm ERR! cwd /home/pi/src/oref0
npm ERR! node -v v0.10.29
npm ERR! npm -v 1.4.21
npm ERR! path /home/pi/src/oref0/node_modules/oref0/bin/bt-pan
npm ERR! code EPERM
npm ERR! errno 50
npm ERR! stack Error: EPERM, chmod '/home/pi/src/oref0/node_modules/oref0/bin/bt-pan'
npm ERR!
npm ERR! Additional logging details can be found in:
npm ERR!     /home/pi/src/oref0/npm-debug.log
npm ERR! not ok code 0

npm ERR! oref0@0.5.0-dev global-install: `npm install && sudo npm install -g && sudo npm link && sudo npm link oref0`
npm ERR! Exit status 50
npm ERR!
npm ERR! Failed at the oref0@0.5.0-dev global-install script.
npm ERR! This is most likely a problem with the oref0 package,
npm ERR! not with npm itself.
npm ERR! Tell the author that this fails on your system:
npm ERR!     npm install && sudo npm install -g && sudo npm link && sudo npm link oref0
npm ERR! You can get their info via:
npm ERR!     npm owner ls oref0
npm ERR! There is likely additional logging output above.
npm ERR! System Linux 4.1.13-v7+
npm ERR! command "/usr/bin/nodejs" "/usr/bin/npm" "run" "global-install"
npm ERR! cwd /home/pi/src/oref0
npm ERR! node -v v0.10.29
npm ERR! npm -v 1.4.21
npm ERR! code ELIFECYCLE
npm WARN This failure might be due to the use of legacy binary "node"
npm WARN For further explanations, please read
/usr/share/doc/nodejs/README.Debian

npm ERR!
npm ERR! Additional logging details can be found in:
npm ERR!     /home/pi/src/oref0/npm-debug.log
npm ERR! not ok code 0
katie disimone
@Kdisimone
May 02 2017 03:24

@scottleibrand does this error exist in the jubilinux? (Since it is kernel?)

May 01 20:21:07 localhost kernel:  [ 1999.742601] wl_host_event: mismatched OUI, bailing
May 01 20:21:08 localhost kernel:  [ 2000.355480] wl_host_event: mismatched OUI, bailing
May 01 20:21:08 lukeedison network.log:  At Mon May  1 20:21:02 PDT 2017, my wifi network name is "ATT2Ajj6Xe", and my public IP is: 104.63.155.106
May 01 20:21:08 lukeedison network.log:  Finished oref0-online.
May 01 20:21:08 localhost kernel:  [ 2000.663114] wl_host_event: mismatched OUI, bailing
May 01 20:21:08 localhost Timed:  out or other comms error - Received an error response Timeout - retrying: 1 of 3
May 01 20:21:09 localhost Timed:  out or other comms error - Received an error response Timeout - retrying: 2 of 3
May 01 20:21:11 localhost Timed:  out or other comms error - Received an error response Timeout - retrying: 3 of 3

This rig has been catching only intermittent loops. It's been runagain several times trying to fix it. I'm thinking it may need a flash though?

Scott Leibrand
@scottleibrand
May 02 2017 03:27
Not sure what that error means. I think OUI is organizational unique identifier, which is probably part of the MAC address. Not sure what a mismatch would mean though. Maybe Google it?
katie disimone
@Kdisimone
May 02 2017 03:28
@esialb any ideas on that error message from kernel (above)
@scottleibrand thanks. Will do!
Scott Leibrand
@scottleibrand
May 02 2017 03:28
@dougtakeuti EPERM is a permissions error. I would remove the oref0 dir and re-clone it.
tynbendad
@tynbendad
May 02 2017 03:49
how long do your phone batteries last when the rig is connecting to it all the time?
my son's died today around 3pm, so about 9 hours
rig lasted fine at least
Tim Street
@tim2000s
May 02 2017 03:57
@tynbendad is he connected via bt or wifi?
katie disimone
@Kdisimone
May 02 2017 04:15
Wifi tether kills phone battery fast. Like by mid day for us. BT tether goes much longer. That's why we setup BT pairing
Lin
@denglfddd
May 02 2017 04:23
@scottleibrand hi,i need your help~thanks
Lin
@denglfddd
May 02 2017 04:53
屏幕快照 2017-05-02 12.53.15.png
tynbendad
@tynbendad
May 02 2017 05:01
@tim2000s bluetooth tether to rig, no wifi at school (none that is worth trying at least)
katie disimone
@Kdisimone
May 02 2017 05:10
@denglfddd double check your interfaces. Post result of cat /etc/network/interfaces
@tynbendad did you enter his hotspot info into your wpa-supplicant list?
tynbendad
@tynbendad
May 02 2017 05:11
no, didn't setup wifi hotspot, if that's what you mean, just bluetooth tether
its an iphone 5c, maybe bluetooth tether is too much for its battery
i have a spare 5s, might try to get him to switch to that
katie disimone
@Kdisimone
May 02 2017 05:12
I thought you'd said you hadn't done BT pairing yet...sorry. My bad.
tynbendad
@tynbendad
May 02 2017 05:14
which phone do you use w/bt tether? can it last all day?
he didn't yell at me about not switching back to wifi today because his phone died right before he got home, so rig switched back ok somehow
Tim Street
@tim2000s
May 02 2017 05:18
How old is the 5c? If it's more than 12 months then I wouldn't expect great things from the battery.
On an android with bt & a 2700mAh battery, I got about 14 hrs life.
katie disimone
@Kdisimone
May 02 2017 05:19
We use an iPhone 6. She gets a full school day on BT tether, but we don't do it often
Tim Street
@tim2000s
May 02 2017 05:20
On android with a 4200 mAh battery I got all day plus some.
I suspect your issue is that the battery in the 5c is past its best.
Combined with the fact that it is rated for about 8.5hrs of internet use with no bt enabled....
Lin
@denglfddd
May 02 2017 05:37
屏幕快照 2017-05-02 13.37.02.png
tynbendad
@tynbendad
May 02 2017 05:37
thanks, guess i'll talk to him about it - he wasn't thrilled with the prospect last time i brought it up :)
Lin
@denglfddd
May 02 2017 05:38
@Kdisimone here
Tim Street
@tim2000s
May 02 2017 05:41
@denglfddd You need to go back to the setting up the Edison documentation. You've not set up /etc/networks/interfaces properly.
JELCRAWFORD
@JELCRAWFORD
May 02 2017 05:48
@tynbendad we put a mophie battery case on Iphone 5s - DS gets through the day w/o killing both phone and mophie battery. Sometimes on wifi, sometimes on BT tether. That said, original battery that came with the phone is on it's last legs - he'll often run through 30% of the battery on the way to school. If we insisted he stay off pokemon go, it would probably go better. He also rolls with a spare battery in his low kit, complete with cord - just a small one to get him through the day if it's going to die towards the end. Occasionally, as the rig battery gets older, he needs to plug that in too, so the small external battery has been useful.
Lin
@denglfddd
May 02 2017 05:48
Can you tell me where to enter the error?
@tim2000s
Tim Street
@tim2000s
May 02 2017 06:00
If you check the docs @denglfddd , you'll see that you are missing the wpa_supplicant setup in your interfaces file. Please go and reread that section of the docs.
Scott Leibrand
@scottleibrand
May 02 2017 06:04
@tim2000s you mean this part?:
auto wlan0
iface wlan0 inet dhcp
    wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf
because his screenshot shows he does have that...
@denglfddd what does your /etc/wpa_supplicant/wpa_supplicant.conf file look like? please don't include the actual network passwords.
Tim Street
@tim2000s
May 02 2017 06:13
My mistake!
Teach me to troubleshoot from a phone.
Cameron Chunn
@applehat
May 02 2017 07:07
@garykidd there's xDrip (for sure Android only) and then Nightwatch (also pretty sure it's Android only). No clue about iPhone. Not even sure how you'd write an iPhone app that does Android auto.
Nightwatch might work on Android 2.0 stand alone tho. Worth looking into.
Android Wear 2.0*
Lin
@denglfddd
May 02 2017 07:32
@scottleibrand i dont know how to check it
Lin
@denglfddd
May 02 2017 07:52
屏幕快照 2017-05-02 15.51.03.png
PieterGit
@PieterGit
May 02 2017 07:57
Did anybody think of using mqtt / mosquito for messaging within a rig, between rigs and with nightscout. i think it might be beneficial to signalling accoss rigs and for notification. Each rig has its own local broker, and tries to forward it to a remote broker (e.g. integrated with nightscout),see e.g. http://www.steves-internet-guide.com/mosquitto-bridge-configuration/ for an example
Topics can be something like:
  • /cgm/loop-start
  • /cgm/sgv
  • /cgm/loop-end
  • /pump/loop-start
  • /pump/loop-end
  • /pump/command/temp-basal
  • /notification/high_alert
  • /notification/low_alert
    Advantages: leight weigt, platform independent, beinig able to have full history in nightscout (even when rigs are offline for several hours), if rigs can connect to the broker they can see if other rigs are doing pump-communication, less cronjobs, pump-loop can start when a new cgm/sgv is available, etc
@denglfddd can you explain what you're trying to accomplish and what doesn't work. you wpa-config looks fine to me on first glance
Lin
@denglfddd
May 02 2017 08:02
@PieterGit I am checking if it is connected to wifi when i enter the following command ifup wlan0,
PieterGit
@PieterGit
May 02 2017 08:07
what does ifconfig wlan0 return?
Lin
@denglfddd
May 02 2017 08:11
屏幕快照 2017-05-02 12.53.15.png
Tim Street
@tim2000s
May 02 2017 08:19
@applehat xDrip plus watchface should work natively with Android Wear 2.0.
@cameronrenwick I am currently running with this as my battery. 4400mah (2 cell version) and it lasts a very long time: https://shop.pimoroni.com/products/lithium-ion-battery-pack
Mike
@libxmike
May 02 2017 09:12
@tim2000s you can't get the panasonic 18650 with 3400mAh in one cell (make sure reliable source like the one in NL or Amazon), I'm designing case for it for Explorer Board but first have to finish my lipo case design.
Lin
@denglfddd
May 02 2017 10:31
@PieterGit
rsilvers129
@rsilvers129
May 02 2017 11:51
Will pass through charging work with two 18650?
Tim Street
@tim2000s
May 02 2017 12:27
As long as they're 3.7v.
Works fine in mine.
Luis Betancourt
@betluis
May 02 2017 12:33
hello all.... I am having a continuous error after Old Pumphistory: Refresh / Error, retrying... does not get over this error continuously. I changed the openaps rig so it is not it. Also nothing seems strange in the pump. any clue?
this is after a bad sensor change last night (Medtronic sensor)
BG continues being updated in NS ok though. the NS pill appears continuously as unkwown
Luis Betancourt
@betluis
May 02 2017 12:44
Old pumphistory: Refresh
Error, retrying
Listening: .No pump comms detected from other rigs
mmtune: pump://JSON/mmtune/monitor/mmtune.json
reporting monitor/mmtune.json
"916.636", 5, -83 Starting pump-loop at Tue May 2 07:39:16 CDT 2017:
Radio ok, Listening: .No pump comms detected from other rigs
MDT cgm data retrieved
MDT cgm data reformatted
Old pumphistory: Refresh
Error, retrying
Listening: .No pump comms detected from other rigs
mmtune: pump://JSON/mmtune/monitor/mmtune.json
reporting monitor/mmtune.json
"916.636", 5, -84 Starting pump-loop at Tue May 2 07:41:58 CDT 2017:
Radio ok, Listening: .No pump comms detected from other rigs
MDT cgm data retrieved
MDT cgm data reformatted
Old pumphistory: Refresh
Error, retrying
Luis Betancourt
@betluis
May 02 2017 12:54
looks like if there was a problem with the pumphistory data, maybe something corrupted?... is there a way to delete the pumphistory data in the pump?
AtibhavSinha
@AtibhavSinha
May 02 2017 12:55

Hi Guys,
I am having a requirement to put color/ small images for/in the data label on bar chart.

But with the present C3 implementation I can provide only text to data label . So just wanted to check if there is any custom implementation available to pass HTML to data label ? Or C3 team has any plan to support this ?

cameronrenwick
@cameronrenwick
May 02 2017 13:00
@tim2000s nice battery! do you have a 3d printed case that you hold that along with the edison/explorer in?
cameronrenwick
@cameronrenwick
May 02 2017 13:23
Thanks @rsilvers129 .. I found a Canadian supplier that has the 4400's in stock so I'm getting 2 on the weekend. Just need to source a case and should be good to go
that's the trouble with the US/Can border... no shipping of lipos over it...
rsilvers129
@rsilvers129
May 02 2017 13:24
I am not sold on them yet. Are they more space efficient than two of the flat 2500s?
rsilvers129
@rsilvers129
May 02 2017 13:31
Two 2500s are 14.25mm thick. 18650s are 18.35mm thick.
I would save 18650s for flashlights, not non-round items designed to be thin.
katie disimone
@Kdisimone
May 02 2017 14:23
@denglfddd I see extra spaces in your wpa_supplicant lines before and after your =
Tim Street
@tim2000s
May 02 2017 14:31
The main reason for using the 18650s is that the are considerably more robust than the Lipos.
cameronrenwick
@cameronrenwick
May 02 2017 14:40
@tim2000s that's my finding too. I've used lipos and the 18650s (granted for other devices) but have better lasting time with the 18650s. Just fiddling now with case designs
Chris Oattes
@cjo20
May 02 2017 14:42
tim might have meant less likely to explode when knocked
Tim Street
@tim2000s
May 02 2017 14:43
Or even slightly grazed with an explorer board...
garykidd
@garykidd
May 02 2017 14:43
SLightly off topic but, one doenst need to root phone in order to run G4 xdrip set-up ...right?
Tim Street
@tim2000s
May 02 2017 14:43
No
garykidd
@garykidd
May 02 2017 14:50
and if I get a g4 dex transmitter---I wont need the stand-alone receiver or wixel if I have an android phone running xdrip?
Savek-CC
@savek-cc
May 02 2017 14:56
for G4 you need a wixel
G5 is able to talk directly via BT
garykidd
@garykidd
May 02 2017 14:59
what's the most reliable in terms of avoiding dropped connections/ missing BG datapoints. @danamlewis what are you using..
Arlene Samowich
@webwiztn
May 02 2017 15:07
@denglfddd Remove the quotation marks around your ssid and psk values!
katie disimone
@Kdisimone
May 02 2017 15:23
@webwiztn and @denglfddd the quotes are supposed to be there.
Arlene Samowich
@webwiztn
May 02 2017 15:27
@Kdisimone @denglfddd Wow. I swear I read to remove the quotes!
katie disimone
@Kdisimone
May 02 2017 15:28
@webwiztn probably thinking about remove the apostrophe from the iPhone names for wifi tethering purposes
As in rename phone to not have an apostrophe
garykidd
@garykidd
May 02 2017 15:39
what hardware /resources is the explorer board missing that's made available with including a wixel?
Dana Lewis
@danamlewis
May 02 2017 16:05
@garykidd I just use a G4 Share receiver
jmcrawford
@jmcrawford
May 02 2017 16:19
@garykidd The wixel radio chip operates at the same frequency as the dex transmitter - different frequency than the explorer operates at.
Matthias
@ceben80
May 02 2017 16:32

@PieterGit I have an issue with my WW pump.
I'am running for several months with my Edison with mini breakout board and ERF.
The edison has been re-flashed and fully new installed with the latest dev oref0-setup.sh today.

The comm via get_rfspy_version.pyis working fine

root@eddie:~/src/subg_rfspy/tools# ./get_rfspy_version.py /dev/ttyMFD1
RileyLink OK
Version: subg_rfspy 0.8
Response: subg_rfspy 0.8

but mmtuneis throwing an exception

serial.serialutil.SerialException: [Errno 2] could not open port =/dev/ttyMFD1: [Errno 2] No such file or directory: '=/dev/ttyMFD1'
root@eddie:~/myopenaps# ls -ls /dev/ttyMF*
0 crw-rw---- 1 root dialout 4, 128 May  2 18:22 /dev/ttyMFD0
0 crw-rw---- 1 root dialout 4, 129 May  2 18:24 /dev/ttyMFD1
0 crw--w---- 1 root tty     4, 130 May  2 18:22 /dev/ttyMFD2

please advise what to do, I have no idea at the moment......

same pump is working fine with an explorer board and the same setup
Matthias
@ceben80
May 02 2017 16:43
the rig was working fine yesterday, that's odd
Due to this behavior I have flashed it with the newest jubilinux from April 2017 and as said new dev installed
Sgoldberg89
@Sgoldberg89
May 02 2017 16:44
Hi friends, I have a strange situation that I am wondering if anyone has had this happen to them as well. My OpenAPS and Pump pills on Nightscout don't show anything but the nightscout is still updating with pump data like boluses, temp basals etc. and OpenAPS seems to be working fine. My loops are pulling down all correct data as well...
I was going to do what I normally do when I run into a problem: reinstall everything and start from scratch, but wanted to see if there was a fix for this before I just hit the destruct button.
Geoff Aumaugher
@geoffro10
May 02 2017 16:47
@Sgoldberg89 I have the same problem right now. Have had it for about 4 days. I've looked at multiple logs and I can't find anything wrong. I'm very close to starting over myself. Hopefully, we can get it figured out before then. FYI, I'm using xDrip+ as the CGM data source.
Sgoldberg89
@Sgoldberg89
May 02 2017 16:48
@geoffro10 Ok, sorry to hear you're in the same boat but nice to know I'm not alone! I use g4 share
Dana Lewis
@danamlewis
May 02 2017 16:49
@Sgoldberg89 @geoffro10 have you both checked for future treatments in NS? Probably not it but worth a check regardless
Geoff Aumaugher
@geoffro10
May 02 2017 16:49
@danamlewis Yep, no future treatments.
Dana Lewis
@danamlewis
May 02 2017 16:49
we've been looking for ppl to help troubleshoot this scenario so if you have some patience please don't reset up ;)
Geoff Aumaugher
@geoffro10
May 02 2017 16:50
Ok, let's get it figured out!! Here to help. :D
Dana Lewis
@danamlewis
May 02 2017 16:50
K. Does your NS-loop have any weirdness in it? (Or tell me some more about what all you've checked)
Sgoldberg89
@Sgoldberg89
May 02 2017 16:52
One thing that I think is probably worth noting, thought it's also likely not important, is that I am running multiple rigs, so I don't think it's an error with a specific rig
*though
Geoff Aumaugher
@geoffro10
May 02 2017 16:53
Here's my ns-loop.log
blob
I, too, have two rigs running.
Matthias
@ceben80
May 02 2017 16:56

@PieterGit same with script from https://github.com/oskarpearson/mmeowlink/wiki/Non-USA-pump-settings and this mmtune

mmtune = ! bash -c "/usr/local/bin/subg-ww-radio-parameters; echo -n \"mmtune: \" && openaps report invoke monitor/mmtune.json; grep -v setFreq monitor/mmtune.json | grep -A2 $(json -a setFreq -f monitor/mmtune.json) | while read line; do echo -n \"$line \"; done"

output from subg-ww-radio-parameters

root@eddie:~/myopenaps# openaps mmtune
+ cd /root/src/subg_rfspy/tools
+ ./reset.py /dev/ttyMFD1
RileyLink OK
Version: subg_rfspy 0.8
OK
+ sleep 2
+ ./change_setting.py /dev/ttyMFD1 0x06 0x00
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x0C 0x59
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x0D 0x66
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x0E 0x33
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x0F 0x62
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x10 0x1A
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x11 0x13
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x09 0x24
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x0A 0x2E
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ ./change_setting.py /dev/ttyMFD1 0x0B 0x38
RileyLink OK
Version: subg_rfspy 0.8
SUCCESS
+ sleep 0.5
+ exit 0

mmtune: pump://JSON/mmtune/monitor/mmtune.json
monitor/mmtune.json  raised  [Errno 2] could not open port =/dev/ttyMFD1: [Errno 2] No such file or directory: '=/dev/ttyMFD1'
Scott Leibrand
@scottleibrand
May 02 2017 16:57
@denglfddd I can't tell for sure, but your " marks might be the problem: yours look angled, and normally the kind we want show up as straight. try copying and pasting instead of typing those.
Dana Lewis
@danamlewis
May 02 2017 16:58
I have a hunch it's related to a wonky piece of data somewhere. Some people (DO NOT DO THIS) blow away all of their devicestatus collection at first sign of a problem. But I think I saw someone delete a day or two worth of device status back until when they had the problem. @geoffro10 @Sgoldberg89 . So if @scottleibrand doesn't pipe in here with some better troubleshooting ideas, that would be what I have ya'll try (knocking off a few days back to when you remember it working)
I will actually cry if you delete it all - so don't do that.
Or if you know what day it broke, try deleting that day first
I'd still like to figure out WHY this occurs though, but if this solution works we can drill down into someone's data and find what bugs it out.
Or maybe @Sgoldberg89 @geoffro10 one of you is willing to look at your device data from day it broke ?
Scott Leibrand
@scottleibrand
May 02 2017 17:00
another thing you could do is make a copy of your devicestatus collection in mongo and create a new fresh one
Geoff Aumaugher
@geoffro10
May 02 2017 17:00
@danamlewis I'm looking at devicestatus.json on my Heroku site and I don't see any entries since 4/29.
Scott Leibrand
@scottleibrand
May 02 2017 17:00
do you see anything odd-looking in the few most recent ones?
Dana Lewis
@danamlewis
May 02 2017 17:01
Oh that's interesting
Geoff Aumaugher
@geoffro10
May 02 2017 17:01
No, I only see 5 entries and it's just mentioning uploader battery only.
Dana Lewis
@danamlewis
May 02 2017 17:02
@Sgoldberg89 did yours also stop back on the day your pill stopped displaying?
@geoffro10 this wasn't a new rig was it? Previously working rig that looped?
Geoff Aumaugher
@geoffro10
May 02 2017 17:04
@danamlewis Yep, 2 rigs.. been running for months.
Sgoldberg89
@Sgoldberg89
May 02 2017 17:04
sorry @danamlewis, trying to figure out how to see the records from back then
Geoff Aumaugher
@geoffro10
May 02 2017 17:05
@scottleibrand I'm trying to view my last records in mLab and it's timing out.
@danamlewis @scottleibrand Ok, I found the last good devicestatus entry and the next one that doesn't have any info. What should I look for?
Sgoldberg89
@Sgoldberg89
May 02 2017 17:08
Looks like mine stopped on the 29th too
Dana Lewis
@danamlewis
May 02 2017 17:09
Q - do you remember anything about that day? Did your rig die (battery drain all the way down for example) or anything?
Sgoldberg89
@Sgoldberg89
May 02 2017 17:10
I was at Camp Leo over the weekend... No cell service in Mt. Rainier Natl Park so my edison was off and my Pi (which is at home) was just running over the weekend without connecting.
Geoff Aumaugher
@geoffro10
May 02 2017 17:10
My last "good" entry was 4/28 at 22:52 eastern time.
Dana Lewis
@danamlewis
May 02 2017 17:11
@geoffro10 which type are your two rigs?
Geoff Aumaugher
@geoffro10
May 02 2017 17:12
Next entry isn't until 4/29 at 8:50. I have two edisons. Both running with xDrip as CGM source.
Sgoldberg89
@Sgoldberg89
May 02 2017 17:13
@geoffro10 can you show me what a successful entry in the log looks like so I can make sure I am looking for the right thing?
Dana Lewis
@danamlewis
May 02 2017 17:13
Hm, k. And after that gap, did it give more good data or the 4/29 stuff is just the battery ones?
Meghan Rutledge
@mar5477
May 02 2017 17:13
Hey all, I think I am having a similar issue? On 4/29 my xdrip+ stopped uploading to NS
Geoff Aumaugher
@geoffro10
May 02 2017 17:15
@danamlewis The 4/29 entry only shows the $oid, uploaderBattery, and created at. Nothing else.
7 lines only.
Dana Lewis
@danamlewis
May 02 2017 17:15
@mar5477 are you able to loop, but OpenAPS pill not working?
Meghan Rutledge
@mar5477
May 02 2017 17:18
I am able to pull info from the pump into NS, but none of the BGs are uploading so it's not actually 'looping'
Dana Lewis
@danamlewis
May 02 2017 17:19
Ok so this might be a separate problem then @mar5477.
Meghan Rutledge
@mar5477
May 02 2017 17:21
Not sure if the issue is with xdrip or mlab, but xdrip logs say "Unable to do REST API download
: unexpected baseURI"
cameronrenwick
@cameronrenwick
May 02 2017 17:25
@mar5477 I've had that happen too.. and have the same setup as you. It generally clears itself..I've come to believe it's on the mongo end (problem-wise) not your end. I too had this same thing a week ago and ended up finding a problem with one entry in my devicestatus log that was messed up. I deleted that log and the ones that occurred after it (ie more recent past) and it seemed to clear up. It's tough manually going thru the device status entries but that's where my prob was
Meghan Rutledge
@mar5477
May 02 2017 17:26
@cameronrenwick I found the last device status entry that wasn't "uploader battery" is that the one I delete or are there other things I should be looking for?
Geoff Aumaugher
@geoffro10
May 02 2017 17:26
@Sgoldberg89 When I look at a good entry in mLab, it shows 110 lines of information. The next entry only shows 7.
Scott Leibrand
@scottleibrand
May 02 2017 17:29
devicestatus is used for both kinds of device status entries. you'll probably want to look at the last few of each type and see if anything is odd
Sgoldberg89
@Sgoldberg89
May 02 2017 17:31
oh, interesting
I don't have any device-status updates in Mongo after 4/25
which is probably when it stopped working
sorry mlab?
Geoff Aumaugher
@geoffro10
May 02 2017 17:34
@scottleibrand I looked back and everything looks good right up to where it stopped. The last full entry in devicestatus looks good.
Eric
@ecc1
May 02 2017 17:35
@PieterGit I've been thinking about that issue, although not mqtt specifically, for my loop code (which isn't openaps, as you know). Now that I can read G4 transmissions directly, I want to decouple the "glucose providers" (i.e. read from G4 receiver vs. read from sensor) from the rest of the loop, with a simple "persistent publishing" mechanism. That would let me decouple the NS uploader too. My first version is actually using sqlite, since I don't need this to work between machines and I don't want to depend on Internet connectivity anywhere critical.
Geoff Aumaugher
@geoffro10
May 02 2017 17:39
@scottleibrand @Sgoldberg89 @danamlewis My devicestatus updates stopped mid day on 4/29.
Dana Lewis
@danamlewis
May 02 2017 17:42
@geoffro10 did you have any xdrip related problems that day? Or rig problems?
Geoff Aumaugher
@geoffro10
May 02 2017 17:44
@danamlewis No, but was working during that time.
Sgoldberg89
@Sgoldberg89
May 02 2017 17:46
Another issue is that IOB does not get updated based on OpenAPS basal adjustments
Only on boluses
Scott Leibrand
@scottleibrand
May 02 2017 17:47
yeah, IOB, purple line predictions, and OpenAPS pill data are all populated from the devicestatus uploads
you might want to drill down into running that upload report manually
and/or try swapping NS over to a new devicestatus collection (without deleting the old one) to see if that fixes it
Geoff Aumaugher
@geoffro10
May 02 2017 17:49
@scottleibrand How do you swap to a new devicdstatus collection?
*devicestatus
Meghan Rutledge
@mar5477
May 02 2017 17:53
@Geoff1313_twitter if you go into your xdrip app and then to System status and swipe from right to left (to see the uploaders tab) what does it say?
I think to swap device status collections log into your mlab, under collections click "add collection" and name it something like "devicestatus2" then in your mongodb settings in xdrip update the "Device Status collection name" to whatever you just created
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 18:03
I am having problems with my rig when I am connected to my phones wifi or bluetooth and somtimes when connected to my work's wifi I am connected to it now via ssh and would love some assistance getting this figured out @scottleibrand has been helping me some here a brief look at the troubleshooting steps can me found on this link https://www.imzy.com/openaps/post/openaps_does_not_seem_to_be_communicating_with_my_
Meghan Rutledge
@mar5477
May 02 2017 18:04
I think there's another step for fixing it on the rig, but I'm not sure what that is.
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 18:04
I can ping google and it is online
I am sure this has something to do with stale BG data.. This is from the pump log and it is the same thing I see on my NS site "The last BG data was read at Mon May 01 2017 16:48:16 GMT-0500 (CDT) but your system time currently is Mon May 01 2017 17:46:00 "
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 18:12
I can ping google from my rig, I can also ssh to it from my phone and laptop
Sgoldberg89
@Sgoldberg89
May 02 2017 18:14
so I tried creating a second device status collection. Just want to make sure I did it correctly. I created a new log called "devicestatus2" in mLab and then, under admin tools in nightscout, under Roles, I changed device status to api:devicestatus2:create
is that correct?
It hasn't fixed anything, so I wanted to make sure
Meghan Rutledge
@mar5477
May 02 2017 18:18
Is It anything posting to your devicestatus2?
Sgoldberg89
@Sgoldberg89
May 02 2017 18:19
negatory
Meghan Rutledge
@mar5477
May 02 2017 18:21
What does your xdrip app>system status> tab say?
*uploaders
Sgoldberg89
@Sgoldberg89
May 02 2017 18:22
i don't [think] I use xdrip
cameronrenwick
@cameronrenwick
May 02 2017 18:22
the other thing you should likely do @Sgoldberg89 is reset your dyno (in heroko under "more") if you haven't already
Sgoldberg89
@Sgoldberg89
May 02 2017 18:23
@cameronrenwick Thank you! I did do this, but appreciate the tip!
Meghan Rutledge
@mar5477
May 02 2017 18:23
@Sgoldberg89 how do you get your bg in NS?
Sgoldberg89
@Sgoldberg89
May 02 2017 18:23
G4 share
Meghan Rutledge
@mar5477
May 02 2017 18:24
Plugged into your Edison?
Sgoldberg89
@Sgoldberg89
May 02 2017 18:25
g4 --(bluetooth)--> iPhone ------> Dexcom Share ----> Magic -----> Nightscout
Dana Lewis
@danamlewis
May 02 2017 18:25
(Lol)
Meghan Rutledge
@mar5477
May 02 2017 18:27
Hmm, I'm not familiar with that set up, but it makes me think the issue is with mlab?
Sgoldberg89
@Sgoldberg89
May 02 2017 18:28
Or something changed in the handshake that takes place between nightscout and mlab?
Geoff Aumaugher
@geoffro10
May 02 2017 18:30
I use G5 -->xDrip+ --> Edison via BT --> Nightscout
jmcrawford
@jmcrawford
May 02 2017 18:31
@geoffro10 Can you check the nightscout/azure logs, is there anything unusual in there?
Geoff Aumaugher
@geoffro10
May 02 2017 18:32
@jmcrawford I looked at the logs today but don't see anything unusual. Although, I'm not necessarily sure what to look for that would be unusual. LOL
FYI, I'm referring to the Heroku logs.
jmcrawford
@jmcrawford
May 02 2017 18:34
I had an issue a while back where NS stopped displaying anything. Nightscout logs (can't remember if it was before or after we switched to heroku) kept complaining about a record missing a Timestamp field. I was able to locate and delete that record in mlab, fixed the problem. Wouldn't have found it without consulting the nightscout logs.
Geoff Aumaugher
@geoffro10
May 02 2017 18:36
@jmcrawford Where do you see the nightscout logs?
Sgoldberg89
@Sgoldberg89
May 02 2017 18:36
papertrail, I think
Scott Leibrand
@scottleibrand
May 02 2017 18:36
this particular set of issues looks less like a nightscout display issue and more like an openaps -> NS API -> mongo upload related thing
jmcrawford
@jmcrawford
May 02 2017 18:36
@ceben80 The error you're seeing looks odd - complaining about '=/dev/ttyMFD1' instead of '/dev/ttyMFD1'. Given that mmtune can find that device/file, we know it exists. I don't know where that string is entered aside from pump.ini, is there a typo there, and extra =?
@geoffro10 In Heroku, there is a "More" button in the upper right corner, with a "view logs" option
garykidd
@garykidd
May 02 2017 18:39
@PieterGit when you say "i.e. read from ...[G4] sensor directly......are you referring through a wixel?....or someo other mechanism? Thx
Geoff Aumaugher
@geoffro10
May 02 2017 18:40
@jmcrawford Ok, that's what I was looking at. I see "OpenAPS hasn't reported a loop yet". Not really sure what else to look for. I don't see errors per se.
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 18:42
what can I do to allow my Open APS to talk to the NTP servers? for whatever reason I am not able to communicate with them.. I am using the internet on my phone and there are not restrictions and I am online.
root@in10aps:~# killall -g openaps; oref0-set-device-clocks
killall: getpgid(17489): No such process
Waiting for ntpd to synchronize... No!
ntpd did not synchronize.
sudo: unable to resolve host in10aps
[ ok ] Restarting ntp (via systemctl): ntp.service.
Waiting for ntpd to synchronize... No!
ntpd did not synchronize.
root@in10aps:~#
Sgoldberg89
@Sgoldberg89
May 02 2017 18:43
Can I copy the bit from my log regarding device status and you guys can tell me if anything is off?
api:devicestatus:read
api:devicestatus:create
got data-received event, requesting reload
Load Complete:
sgvs:395, treatments:352, profiles:1, tempbasalTreatments:270, tempTargetTreatments:1
garykidd
@garykidd
May 02 2017 18:44
@geoffro10 ....your set up of....I use G5 -->xDrip+ --> Edison via BT --> Nightscout.....how consistent/reliable would you say the uploads of BGs are
?
jmcrawford
@jmcrawford
May 02 2017 18:45
@Sgoldberg89 You're not at your data capacity limit in mlab, are you?
Sgoldberg89
@Sgoldberg89
May 02 2017 18:45
no, only 35mb
wait
no, it says it's up and running
Geoff Aumaugher
@geoffro10
May 02 2017 18:48
@garykidd They are very consistent. Works great for using whatever data connection my phone has.
Matthias
@ceben80
May 02 2017 18:50
@jmcrawford fantastic, two eyes more....thank you so much, it was really a typo, my bad :worried:
@PieterGit forget my question/post :wink:
@scottleibrand ....murphy today, it's working
garykidd
@garykidd
May 02 2017 18:50
so the g5 transmitter connects to android phone directly which is running xdrip+?
Geoff Aumaugher
@geoffro10
May 02 2017 18:52
Yes... G5 connects to xdrip+ via Bluetooth on my Pixel phone.
garykidd
@garykidd
May 02 2017 18:54
and your edison rig connects to your pixel via BT as well?
Geoff Aumaugher
@geoffro10
May 02 2017 18:55
Yes
Dana Lewis
@danamlewis
May 02 2017 18:55
@ceben80 it's nice when it's as easy as a typo fix! Props @jmcrawford for the spot.
garykidd
@garykidd
May 02 2017 18:55
Did you happen to ever try G5 via iphone
@geoffro10
Sgoldberg89
@Sgoldberg89
May 02 2017 18:56
logging off now, let me know if you guys figure out what's going on
Geoff Aumaugher
@geoffro10
May 02 2017 19:02
@garykidd Yes, used iPhone for a while with Dexcom software until I discovered xDrip and realized I could go back to Android.
Matthias
@ceben80
May 02 2017 19:03
@danamlewis one positiv aspect, the rig has new jubi flashed, last version :smile:
garykidd
@garykidd
May 02 2017 19:05
I had had fairly regular drops of BG readings via the stadard G5 mobile / iphone setup....more recently it actually seemed a lot better....but still seriously considering switching to android setup----do you think it's more reliable overall---or do you just prefer android?
Geoff Aumaugher
@geoffro10
May 02 2017 19:06
I think the reliability has been comparable. I prefer android so that's why I switched.
Eric
@ecc1
May 02 2017 19:06
@diabetic_dude_twitter I got rid of NTP on my rigs a while ago and am using systemd's built-in time sync service. It's been working fine for me. To change over:
$ sudo apt-get purge ntp
$ sudo systemctl enable systemd-timesyncd
$ sudo systemctl start systemd-timesyncd
$ sudo timedatectl set-ntp yes
(Edited to actually start the service, if you're not rebooting)
BDavis1960
@BDavis1960
May 02 2017 19:15
Pump loop 5:2.png
Hi all. Sorry for another SOS -I hope to be on the other end assisting someday. Had a great first looping experience over the weekend over wifi and BT. However, when I returned home on Sunday nite, my rig did not reconnect, and I have not been able to fix. Here are some logs. I see "old pump history, error, retrying" Is that the prob? Appreciate any advice.
Network script 5:2.png
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 19:16

```
root@in10aps:~# systemctl start systemd-timesyncd
Job for systemd-timesyncd.service failed. See 'systemctl status systemd-timesyncd.service' and 'journalctl -xn' for details.
root@in10aps:~#

``` @ecc1

augh close enough
Eric
@ecc1
May 02 2017 19:26
And what does the log say?
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 19:27
systemctl status systemd-timesyncd.service said essentially failed
root@in10aps:~# systemctl status systemd-timesyncd.service
● systemd-timesyncd.service - Network Time Synchronization
   Loaded: loaded (/lib/systemd/system/systemd-timesyncd.service; enabled)
   Active: failed (Result: start-limit) since Tue 2017-05-02 14:14:09 CDT; 3min 59s ago
     Docs: man:systemd-timesyncd.service(8)
  Process: 5125 ExecStart=/lib/systemd/systemd-timesyncd (code=exited, status=226/NAMESPACE)
 Main PID: 5125 (code=exited, status=226/NAMESPACE)

May 02 14:14:09 in10aps systemd[1]: Unit systemd-timesyncd.service entered ...e.
May 02 14:14:09 in10aps systemd[1]: systemd-timesyncd.service has no holdof...t.
May 02 14:14:09 in10aps systemd[1]: Stopping Network Time Synchronization...
May 02 14:14:09 in10aps systemd[1]: Starting Network Time Synchronization...
May 02 14:14:09 in10aps systemd[1]: systemd-timesyncd.service start request...t.
May 02 14:14:09 in10aps systemd[1]: Failed to start Network Time Synchroniz...n.
May 02 14:14:09 in10aps systemd[1]: Unit systemd-timesyncd.service entered ...e.
Hint: Some lines were ellipsized, use -l to show in full.
root@in10aps:~# systemctl start systemd-timesyncd
Eric
@ecc1
May 02 2017 19:30
Is there anything relevant in syslog?
Diabetic_Dude
@diabetic_dude_twitter
May 02 2017 19:30
let me see
ok not sure weather to say woohoo or crap
the logs say it is up and it does seem to be up now
exit
oops
garykidd
@garykidd
May 02 2017 19:35
@danamlewis do you think the g4 via receiver or wixel is more reliable than G5 has been reported to be...?
Tim Street
@tim2000s
May 02 2017 19:38
I think it's horses for courses on that one, and your phone makes a big difference on the G5. I've found that G5/x:Drip on most of the set-ups I've had has been reliable.
Dana Lewis
@danamlewis
May 02 2017 19:38
I have no personal experience w G5, but have seen several people with G5s complain. Mostly iPhone users though, so not sure how it does w xdrip
garykidd
@garykidd
May 02 2017 19:39
and g4 doesnt use BT as RF comm correct?
jmcrawford
@jmcrawford
May 02 2017 19:50
We've had some reliability issues with G4/Wixel. Gained much stability when we switched phones, but still had drops. Noticed that drops occurred much more frequently at home, figured something was causing rf interference. I've since put wixel/pi setups around the house, now much more consistent data. Outside the house, the G4/wixel seems quite stable. I think phone you choose plays a big role as well.
@sulkaharo mentioned improvements with using HM-17 bluetooth modules instead of HM-10/11
alternateal
@alternateal
May 02 2017 20:15
I have found G5 since latest OS update and Dexcom app update to be massively more stable than before. I don't think I have had a failure since the last update
garykidd
@garykidd
May 02 2017 20:18
@alternateal nice to here---I have had better results too so that's encouraging that we've both seemed to expereience an improvement....although still not perfect.
alternateal
@alternateal
May 02 2017 20:19
Pancreabble question - I am trying to use offline from Edison. I followed the latest PR instructions, and I can get the notifications on bottom of urchin screen (IOB etc) but can't get the CGM data or graph - am I missing something simple like a setting in urchin??
garykidd
@garykidd
May 02 2017 20:22
and is dexcom planning on support for andoid soon as well for G5?
Eric
@ecc1
May 02 2017 20:22
@garykidd correct, G4 transmitters use proprietary 2.4GHz comms (that's what Wixels are used for if you're not using a G4 receiver)
garykidd
@garykidd
May 02 2017 20:23
@ecc1 that's what I thought too---doubted myself when I saw a recent post about g4 using BT recently here (perhaps/liklely I just misunderstood).
Eric
@ecc1
May 02 2017 20:25
the xdrip hardware has both Wixel and BT; Wixel to receive the sensor data, BT to relay it to the xdrip app on the phone (or directly to Edison as some are doing)
garykidd
@garykidd
May 02 2017 20:25
@ecc1 I was thinking that it might be worth reverting to G4/wixel or G4/share receiver vs. G5/BT/Iphone....given that's its altogether a different protocol
I'm sure it'd be a bunch of work and not altogether neccesary but fairly sure somone thought about consolidating wixel component on to an edison---seems like certain resource duplication might be eliminated......2 batteries for 1.....2 BT modules for 1....etc.
sorry not edison meant explorer as an example...
Eric
@ecc1
May 02 2017 20:30
I'm one of those someones :-) I currently have it running in prototype mode, not out and about with me on my rig, while I tweak the calibration algorithm.
Actually using radio-only CC2500, versus the wixel which has radio + MCU
garykidd
@garykidd
May 02 2017 20:31
very interesting......
and its not possible to use Dex's calibration methodology ....I'm guessing that part of the secret wasnt revealed?
Eric
@ecc1
May 02 2017 20:35
The xdrip app does its own calibration, but I don't quite understand it yet
then there's the UI issue of how to tell the Edison about the fingersticks (for openaps the obvious way would be NS careportal I guess, but I don't want to rely on NS)
garykidd
@garykidd
May 02 2017 20:38
hmmm....could use just use the pump to manually input BG reading perhaps? you'd be realying on the rig picking up the imput...but not necessarily the rig uploading to NS I guess...
Eric
@ecc1
May 02 2017 20:39
yes, that's a reasonable idea
since the rig is already getting that history from the pump
garykidd
@garykidd
May 02 2017 20:42
hmmm....I dont suppose MM will just cave and switch all RF comms to BT
and release all the data we need to manipulate it...
haa haa
alternateal
@alternateal
May 02 2017 20:43
@garykidd - thought the 630g onwards was all BT only? You are right on your latter points however :worried:
Martin Haeberli
@mhaeberli
May 02 2017 20:45
@alternateal - from what I’ve seen in the 630g / 670g NightScout channels, the 630g/670g are NOT BT only
but wait
i do know they use 2.4 ghz communications
and current nightscout efforts use the updated link meter.
since BT is also 2.4 ghz, it is possible...
alternateal
@alternateal
May 02 2017 20:53
@mhaeberli - I guess calling it NextLink 2.4 is a bit of a give away - my assumption that it is Bluetooth given the latest smartphone app for Medtronic CGM.
Sgoldberg89
@Sgoldberg89
May 02 2017 20:57
@danamlewis do you think I should create a new nightscout?
Martin Haeberli
@mhaeberli
May 02 2017 21:07
on another node - one of my Edisons won’t flash - i’m getting “ Ready to receive application ” on the serial / uart port, but no joy getting flash to happen. I guess I’ll have to try recovery
Scott Leibrand
@scottleibrand
May 02 2017 21:09
did you swap cables to confirm they both work as data cables?
Meghan Rutledge
@mar5477
May 02 2017 21:10
@Sgoldberg89
Tim Street
@tim2000s
May 02 2017 21:10
640G and 670G use Zigbee rather than BT
Martin Haeberli
@mhaeberli
May 02 2017 21:10
yes
that said, swapping again
Meghan Rutledge
@mar5477
May 02 2017 21:11
Trying to figure out if we are having similar issues, can you walk me through it again.
Sgoldberg89
@Sgoldberg89
May 02 2017 21:14
@mar5477 Nightscout OpenAPS pill reads "unknown" and won't update. Meanwhile the loop is still running
Martin Haeberli
@mhaeberli
May 02 2017 21:15
@tim2000s thx for that note re Zigbee
Tim Street
@tim2000s
May 02 2017 21:21
np
Sgoldberg89
@Sgoldberg89
May 02 2017 21:26
@mar5477 @danamlewis @geoffro10 , I noticed over the past week that I've been going v high after meals and I think it's because there's a failure in the loop to process COB. I think what's happening is it's looking at IOB and when it bumps into the Max IOB setting in the loop, it cuts my basal way down. So it isn't an issue overnight but it means I'm getting a reduced basal immediately after eating
I think it's a bigger problem then I was aware of before
*than
Scott Leibrand
@scottleibrand
May 02 2017 21:28
how do you enter carbs?
Sgoldberg89
@Sgoldberg89
May 02 2017 21:28
bolus wizard?
Scott Leibrand
@scottleibrand
May 02 2017 21:28
oref0 shouldn't rely on nightscout to read carbs then
you can look at pump-loop.log to see how many COB it thinks you have
Sgoldberg89
@Sgoldberg89
May 02 2017 21:31
I do see the COB in the log. I just don't recall the loop dropping basal rates below standard when bg is high and raising.
Scott Leibrand
@scottleibrand
May 02 2017 21:36
find an example where it's setting low temps when you think it should be high temping, an let's look at the pump-loop.log for that run.
PieterGit
@PieterGit
May 02 2017 21:44
@ecc1: i'm thinking of a local mqtt broker on the rig (perhaps even bound to 127.0.0.1 for security), so no internet connectivity is required if you have one rig. The uploading to nightscout might be used with an internet reachable mqtt broker which the rigs forward the information to (when online), e.g. https://devcenter.heroku.com/articles/cloudmqtt (no Go bindings described :smile:, but i'm sure there are mosquitto bindings for Go ). Because of firewalls/NAT etc. the rigs should forward it to the internet somehow. Then Nightscout can read the topics from the internet reachable mosquitto instance.
@denglfddd did removing the extra space in wpa_supplicant lines fix your error? I couldn't find a reaction from you (but maybe i missed it)
@Kdisimone: you shouldn't post wifi passwords online, please remove these from gitter (and change your passwords)
katie disimone
@Kdisimone
May 02 2017 21:52
@PieterGit I take the screenshots for the express purpose of of posting in the docs publicly. That screenshot was taken from the docs. These are not my actual passwords. But thanks!
PieterGit
@PieterGit
May 02 2017 21:53
ok
i don't need the docs to setup wifi anymore :smile:
Meghan Rutledge
@mar5477
May 02 2017 22:27
More weird things are happening over here. Heroku logs say 2017-05-02T21:38:39.170Z EMITTING ALARM:
2017-05-02T21:38:39.171008+00:00 app[web.1]: {"level":"Urgent","title":"URGENT: Pump Battery Low","message":"Pump Battery: 1.29v\nPump Reservoir: 138.9U","group":"Pump","plugin":"pump"} but nothing is happening on my NS site
susieanneh
@susieanneh
May 02 2017 22:47
Hey guys! I'm going to get started soon. I have all the hardware. I was wondering how long roughly it would take with little experience. I am wondering if I can get this done before my toddlers class is over. About 3 hours. Or if I need more time. Thanks!
Dana Lewis
@danamlewis
May 02 2017 22:48
@susieanneh is your Edison pre-flashed?
with a pre-flashed Edison, definitely doable
without a pre-flashed Edison - depends on how easy your flash process is (seems to be based on the computer, not people’s skill)
susieanneh
@susieanneh
May 02 2017 22:53
Yes pre flashed! Thanks! I'm so excited!!!
Dana Lewis
@danamlewis
May 02 2017 22:53
:+1: :)
also don’t hesitate to ask questions if you get stuck or not sure what to do next!
Martin Haeberli
@mhaeberli
May 02 2017 22:55
reiterating - two of three Edisons happily flashed. one is stuck - on *** Ready to receive application ***
advice sought
Geoff Aumaugher
@geoffro10
May 02 2017 22:57
@scottleibrand You said you thought our problems could be an OpenAPS->NS API -> Mongo issue. Anything I can do to check and verify that?
Scott Leibrand
@scottleibrand
May 02 2017 22:59
@geoffro10 to really drill down, you're going to need to try to manually do the upload commands and try to figure out why they're failing
or just start swapping out components until you find the thing that makes it work.
Sgoldberg89
@Sgoldberg89
May 02 2017 23:07
@danamlewis OpenAPS pill logged something 45 min ago but then stopped again
Martin Haeberli
@mhaeberli
May 02 2017 23:07
with intel flash lite tool - device is “flapping” between “detected” and “lost"
Meghan Rutledge
@mar5477
May 02 2017 23:08
@scottleibrand I was able to upload by following the "Offline looping instructions" but used the IP address of my home wifi. If given the option to have multiple REST APIs, why isn't it default to just upload to the rig on whatever wifi it is on?
Scott Leibrand
@scottleibrand
May 02 2017 23:10
@mhaeberli sounds like a flaky cable or USB port. what all have you tried there?
@mar5477 I haven't been involved in any of the xdrip stuff, except to merge PRs.
Sgoldberg89
@Sgoldberg89
May 02 2017 23:11
@danamlewis Looking at the mLab log, the last report was while the pump was suspended, I deleted that one log and then when I suspended it today, it recorded one devicestatus update.
Martin Haeberli
@mhaeberli
May 02 2017 23:13
@scottleibrand using the same explorer board, same cables, same computer on which I was happy successful earlier today. just tried re-seating edison on explorer board and re-seating cables. same thing - amazingly, flash lite kept on trying until about 42% - afraid this edison is probably toast. so it goes
also tried external battery attached as i had seen a rumor someplace that this could affect it
Scott Leibrand
@scottleibrand
May 02 2017 23:15
I would try new cables first.
Dana Lewis
@danamlewis
May 02 2017 23:16
@Sgoldberg89 anything since then? Or just one?
Sgoldberg89
@Sgoldberg89
May 02 2017 23:16
Just the one
I was going to delete it again and suspend to see if it records another
Martin Haeberli
@mhaeberli
May 02 2017 23:17
@scottleibrand thx - i’ll keep group posted … looking a bit more promising...
Sgoldberg89
@Sgoldberg89
May 02 2017 23:17
but wanted to make sure there wasn't any info from it you guys would want to see before I did
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:23
@mhaeberli @alternateal - we have a shiny new 630g sitting on the desk, unfortunately can not loop with it - boo. I can not find anything helpful in the documentation about communication protocols at all - other than your aforementioned 2.4Ghz communications. Pump came with two new Contour Next Link 2.4 meters. I should pop over to your referenced channels to see whats up. Would love to be able to loop with it - I'm sure we are not the only ones in this regard! ps - I mean the Medtronic/ascensia documentation...
Martin Haeberli
@mhaeberli
May 02 2017 23:28
@JELCRAWFORD - definitely possible to get BG values INTO NightScout from 670G and 630G - basically you need to connect one of the new Next Link meters via USB to, say, an Android phone.
two or three more steps: a) control the pump (some hints that the pump has at least remote bolus and I HOPE it has remote basal adjust); b) plug the Giant Next Link into the TINY Edison; c) maybe figure out how to eliminate the new Next Link with a Zigbee radio of some kind … (none of these are simple)
@scottleibrand successful re-flashing edison to stock 201606 Intel release; now about to try jubilinux ...
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:32
@mhaeberli - pump definitely has remote bolus - I've played with it - uses meter to deliver - seems to function as I'd expect (nice feature). Boy has not used it yet... aforementioned looping issue. Maybe to diabetes camp as he won't take his loop there anyhow. No cell service there and no one who will have any idea how to work it...
unfortunately technical issues are not my forte... so not simple for me is probably impossible to solve...
so I'm hoping for some of you really smart people :)
Dana Lewis
@danamlewis
May 02 2017 23:34
@Sgoldberg89 nah, worth trying
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:34
@mhaeberli - Oh yeah - I don't think there is any remote basal function - not that I can see, anyhow... only bolus...
Martin Haeberli
@mhaeberli
May 02 2017 23:35
grrr
as to remote basal
@scottleibrand declaring victory. it was messy, but it is flashing jubilinux now
Scott Leibrand
@scottleibrand
May 02 2017 23:38
there was a significant effort toward remote controlling the 640g, and it basically got to the point of deciding it wasn't worth it, because there's no way to safely suspend insulin delivery temporarily. you could rely on the smartguard feature, but that is a not very good solution.
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:39
@scottleibrand - boo to no way to control...
Martin Haeberli
@mhaeberli
May 02 2017 23:39
@scottleibrand one horribly dumb 630/640/670 g idea - set the main basal at 0.05 U / hour, say - effectively de minimus. Then do a chain of remote boluses, once every 6 - 12 minutes, to simulate basal… probably a BAD idea. But remain bummed that Medtronic isnt a bit more open
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:40
ditto to that
Martin Haeberli
@mhaeberli
May 02 2017 23:40
probably my idea of the biology of repeat boluses is beyond naive ...
repeat (micro) boluses, that is
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:41
It's better than my best idea - find a motivated teenager to break the communication protocal to do what we want... that's it - all I've got... also problematic for so many reasons :)
Dana Lewis
@danamlewis
May 02 2017 23:41
that’s what the ping does. however, huge risk of DKA if you aren’t able to communicate to the rig and you don’t have underlying basals to fall back to .
Martin Haeberli
@mhaeberli
May 02 2017 23:42
@danamlewis thx
Dana Lewis
@danamlewis
May 02 2017 23:43
(so possible, yes, but requires a huge burden on someone to monitor and maintain and keep the system running, which goes above and beyond of what most of us want to do)
JELCRAWFORD
@JELCRAWFORD
May 02 2017 23:44
I suppose one could use the untethered approach to 'manage' the DKA issue, but I'm thinking the chained minibolus would be very tedious and the system would lack so many features...
Martin Haeberli
@mhaeberli
May 02 2017 23:49
@danamlewis of course, still monitoring required w/openaps today - no criticism, but as we try to ramp to full on loop - lots can go wrong - for example, CGM error, CGM data drops, pump communication fails; openaps battery fails - all can result in a low BG suspend failing to happen… so monitoring is alas still the watchword even w/openaps … I love the potential open-ness of the openaps approach, even while drowning in complexity and intended but not yet done PRs … so please hear this as reflective thankfulness!
Sgoldberg89
@Sgoldberg89
May 02 2017 23:49
@danamlewis @geoffro10, Good news, I just deleted the last 3 or 4 logs prior to the issue and things seem to be working again!
absolutely no clue why.
Dana Lewis
@danamlewis
May 02 2017 23:50
@mhaeberli :+1: no worries - love discussions like this, sometimes they help us all push the boundary of what we previously thought possible.
@Sgoldberg89 hahaha. cool. thanks for not blowing away your entire repo, glad it’s working. so the last record was suspend-related?
@geoffro10 before your battery uploads, what behavior was your record related to?
Sgoldberg89
@Sgoldberg89
May 02 2017 23:53
It was, both times. But the first time I only deleted the last record, the second time I deleted 3 or 4
Dana Lewis
@danamlewis
May 02 2017 23:54
do you remember what the other 3-4 were?
Sgoldberg89
@Sgoldberg89
May 02 2017 23:54
I didn't look very closely but they were suspended and they seemed normal otherwise
clearly something was off
Dana Lewis
@danamlewis
May 02 2017 23:54
all were suspend-related?
Sgoldberg89
@Sgoldberg89
May 02 2017 23:55
no, just the very last one
Dana Lewis
@danamlewis
May 02 2017 23:56
k
will be curious to see which behavior the last few of @geoffro10’s records were.
if you see it again, please do the same digging and maybe make additional notes about what the previous records were, and delete one at a time from the problem spot back to the one that was the problem, and that might help us pin down where the root of the problem was. (i.e. something we’re uploading weirdly, or something that just got muffed up and mongo took objection to)
Geoff Aumaugher
@geoffro10
May 02 2017 23:57
@Sgoldberg89 Are you talking about devicestatus records?