502 Bad Gateway

[Elie]

Hi,
So i managed to get rid of 500 error (SMTP) somehow but now when i am trying to register a new user after a minute or two it shows me 502 bad gateway.

I have tried to look into run.py --logs and it shows nothing just no useful info but i will paste an image, any idea what is the bad gateway error ?

[Kal_Lam]

Hi @Elie,

Would you mind having a look at some of the post that has been discussed earlier on the same issue. It should help solve your issue:

Got 502 Bad Gateway accessing url /me

Hello @carrpirates, There are few fixes on /me endpoint within next release of kpi. It’s not ready yet but should be released soon. Meanwhile, you can try to activate the debug mode in kpi and see if you get an error for that. To activate the debug mode, you need to change in kobo-deployments/envfiles/kpi.txt, DEBUG=False to DEBUG=True and restart the KoboInstall. You can read the logs with python run.py --logs

Got 502 Bad Gateway KPI Kobocat

Hi @luisbauti92, Do you see something revelant in KPI and KoboCat logs (kobo-docker/log/kpi/uwsgi.log, kobo-docker/log/kobocat/uwsgi.log) If you have a sentry account (you can create one for free), please set it up with Kobo-Install and paste the error you receive.

Kobo-install gives "502 Bad Gateway"

Hello @nbtl2cu , It seems that polyfill.io was not responding when you tried to install. enketo_express_1 | >> curl: (6) Could no enketo_express_1 | >> t resolve host: polyfill.io enketo_express_1 | Warning: Done, with errors: command “mkdir -p public/js/build && curl “https://polyfill.io/v3/polyfill.min.js?ua=ie%2F11.0.0&features=es2015%2Ces2016%2Ces2017%2Ces2018%2Cdefault-3.6%2Cfetch%2CNodeList.prototype.forEach” -o “public/js/build/ie11-polyfill.min.js”&&cp -f node_modules/enketo-core/src/j…

Have a great day!

[Elie]

Hi @Kal_Lam

I will try to enable debugging and see if anything shows.but where do i look for the debugging log of the 502 ?

[Kal_Lam]

Hi @Elie,

Have a great day!

[Elie]

is this anyhow related?

[Elie]

bumping for any help.

[Elie]

i found this error in access.log
172.25.0.8 - 7c7a72da73fb5aaf132815efbeca2cf37768b760cbe2d5e9d693e8fc949f422eb72ec9333f875b5be9e498788900672b0fb099f075aa5ee9c325d34c [23/Jun/2020:11:35:49 +0000] “POST /api/v2/survey/all HTTP/1.1” 404 134 “-” “python-requests/2.22.0”
172.25.0.7 - - [23/Jun/2020:11:35:52 +0000] “GET /api/v2/asset_snapshots/sEKqZdvNUkgGrppQxteidJ.xml HTTP/1.1” 404 162 “-” “-”

the ip 172.25.x.x is currently used by my network, is it possible to alter the ip ?

[nolive]

You can pull latest version of kobo-install and not exposed backend ports.

[Elie]

i am already on latest version without exposed ports.
i have created a daemon.js in /etc/docker/ and tried to use another subnet but without any success i was looking in kobo-docker for errors in log and i found those errors in kpi.errror.log

[jparmar]

Hey Elie
Did you fix this?
im getting a “1626 upstream timed out (110: Operation timed out) while reading response header from upstream”

[Elie]

no i have not yet, it seems whenever i use port 25 for mail config it does a 500 error whenever i use port 587 with tls i get 502 bad gateway.
where did u get the error from ? which log file?

[jparmar]

i have installed installed portainer.oi and saw it under the nginxcertbot_nginx.

[Elie]

ah u used ssl? i didnt.

[jparmar]

no

[Elie]

i suggest u take a look at iptables or ip route and check if anything is interfering with your private network since it creates virtual adapters and assign private ip’s.

[Elie]

@jparmar

1626 upstream timed out (110: Operation timed out)

on another note i just read, the nginx is probably timing out, try to increate the timeout.

[jparmar]

yes i saw that as well. - not sure its got anything to do with ip etc as its brand new setup.
can i edit the time out in the text file ? or do i need to do it via command?