[9.1.0 On-Premise] AssertionError: cannot be changed at runtime because it is configured on disk

[zoltanlaczko]

Important Details

How are you running Sentry?

• On-Premise docker [Version 9.1.0]
• Saas (sentry.io)
• Other [briefly describe your environment]

Description

Fresh docker install gives internal server error on welcome page. cannot bypass.

Steps to Reproduce

1. Follow the how-to on https://hub.docker.com/_/sentry/ and start sentry first time
2. Create a super-user during upgrade process
3. Login with the super-user
4. On "Welcome to Sentry" page fill the "Root URL" and "Admin Email" and press Continue
5. "Unable to save your changes. Please ensure all fields are valid and try again."

Good items to include here include:

in chrome developer tools the request triggered an Internal Server Error.

in docker logs:
AssertionError: u'mail.port' cannot be changed at runtime because it is configured on disk, 20:40:05 [INFO] sentry.superuser: superuser.request (user_id=1 url=u'http://127.0.0.1:9000/api/0/internal/options/?query=is:required' ip_address=u'xxx.xxx.xxx.xxx' method=u'PUT'), File "/usr/local/lib/python2.7/site-packages/sentry/options/manager.py", line 90, in set, ), '%r cannot be changed at runtime because it is configured on disk' % key, options.set(k, v), File "/usr/local/lib/python2.7/site-packages/sentry/api/endpoints/system_options.py", line 79, in put, response = handler(request, *args, **kwargs), File "/usr/local/lib/python2.7/site-packages/sentry/api/base.py", line 190, in dispatch, response = super(Endpoint, self).handle_exception(exc), File "/usr/local/lib/python2.7/site-packages/sentry/api/base.py", line 90, in handle_exception, Traceback (most recent call last):
default config.yml file (only comments), only the secret_key is set in docker command.

What you expected to happen

Save the configuration without any errors

Possible Solution

[yufuluo]

I tried to deploy 9.1.0 on Kubernetes with helm and got the exact same error, try with default sentry.conf.py and config.yml but no luck.

Here's what I got from api/0/internal/options/?query=is:required:

{"mail.use-tls": {"field": {"disabledReason": null, "default": false, "required": true, "disabled": false, "allowEmpty": true, "isSet": true}, "value": false}, "mail.username": {"field": {"disabledReason": null, "default": "", "required": true, "disabled": false, "allowEmpty": true, "isSet": true}, "value": ""}, "mail.port": {"field": {"disabledReason": "diskPriority", "default": 25, "required": true, "disabled": true, "allowEmpty": false, "isSet": true}, "value": 25}, "system.admin-email": {"field": {"disabledReason": null, "default": "", "required": true, "disabled": false, "allowEmpty": false, "isSet": false}, "value": ""}, "mail.password": {"field": {"disabledReason": null, "default": "", "required": true, "disabled": false, "allowEmpty": true, "isSet": true}, "value": ""}, "mail.from": {"field": {"disabledReason": "diskPriority", "default": "root@localhost", "required": true, "disabled": true, "allowEmpty": false, "isSet": true}, "value": "noreply@company.com"}, "system.url-prefix": {"field": {"disabledReason": null, "default": "", "required": true, "disabled": false, "allowEmpty": false, "isSet": false}, "value": ""}, "auth.allow-registration": {"field": {"disabledReason": null, "default": false, "required": true, "disabled": false, "allowEmpty": true, "isSet": false}, "value": false}, "beacon.anonymous": {"field": {"disabledReason": null, "default": false, "required": true, "disabled": false, "allowEmpty": true, "isSet": false}, "value": false}, "mail.host": {"field": {"disabledReason": "diskPriority", "default": "localhost", "required": true, "disabled": true, "allowEmpty": false, "isSet": true}, "value": "mailhost.company.company.com"}}

[dotconnor]

Currently in the same boat.

Tried manually configuring each available option in config.yml:

auth.allow-registration: false
beacon.anonymous: true
mail.from: ""
mail.host: ""
mail.password: ""
mail.port: 465
mail.use-tls: true
mail.username: ""
system.admin-email: ""
system.url-prefix: ""

However, the onboarding page still shows up

EDIT:
After waiting an hour and then reloading the page, the config settings seemed to have kicked in, and now I'm sent directly to the issues stream. Maybe a cache issue there?

[zoltanlaczko]

confirmed: save the same in config.yml, restart the docker containers and you are in.

[ne1000]

@zoltanlaczko @dotconnor I also met the same issues, save the same in config.yml then docker-compose restart , but I still met the issues

[zoltanlaczko]

@ne1000 : docker stop&rm&run. you have to delete the container before

[dotconnor]

After setting your config in config.yml, if you still got the Setup Sentry page, you need to set the version in the config:

sentry/src/sentry/api/endpoints/system_options.py

Lines 91 to 93 in ea073b1

	# TODO(dcramer): this has nothing to do with configuring options and
	# should not be set here
	options.set('sentry:version-configured', sentry.get_version())

If you're using docker, then run the following command on the web container:

docker exec CONTAINER sentry config set sentry:version-configured '9.1.0'

[ne1000]

@dotconnor yes, it's ok, i'm logined , but how to set up mail delivery ?

[ne1000]

@zoltanlaczko thanks, but it's not ok for me, need to run command as below :
docker exec CONTAINER sentry config set sentry:version-configured '9.1.0'
then it's ok...

[DaveIW2034]

```shell
docker exec CONTAINER sentry config set sentry:version-configured '9.1.0'

thanks @ @dotconnor :))

[zachary95]

If you're modifying your config.yml just like suggested above but nothing happens and you're about to give up on life, you need to rebuild the image.

To sum up :

• Do the edits on config.yml
• Run make build
• Execute docker exec <your_container_name || 'sentry-web-01'> sentry config set sentry:version-configured '9.1.0'
• Then login and tada! 🎉

[ne1000]

@dotconnor Mail still can't be sent. I use SSL protocol instead of tls. Is that the reason?

[itpimp]

Tried all, can logon, but not past the setup page.

[markotitel]

@itpimp Here are the steps that are working.
https://gist.github.com/markotitel/2cbb40ed5a745f0cf3f8ce12395efc11

[Trick93]

@dotconnor
@ne1000
i try run
docker exec onpremise_web_1 sentry config set sentry:version-configured '9.1.0'
i got

04:05:08 [WARNING] sentry.utils.geo: GeoIP module not available.
04:05:08 [WARNING] sentry.utils.geo: No GeoIP MMDB database configured
04:05:12 [INFO] sentry.plugins.github: apps-not-configured
Unhandled exception in thread started by <bound method Thread.__bootstrap of <Thread(Thread-1, stopped daemon 140474876421888)>>
Traceback (most recent call last):
  File "/usr/local/lib/python2.7/threading.py", line 774, in __bootstrap
    self.__bootstrap_inner()
  File "/usr/local/lib/python2.7/threading.py", line 814, in __bootstrap_inner
    (self.name, _format_exc()))
  File "/usr/local/lib/python2.7/traceback.py", line 241, in format_exc
    etype, value, tb = sys.exc_info()
AttributeError: 'NoneType' object has no attribute 'exc_info'

and it also in setup page.

---

update:

i wait a few hours，and it skip the setup pages

[Zegorax]

Hey! What I did was to set up Sentry with v9.0.0, and then upgrade it to 9.1.0. The issue won't occur.

…

Le 18 avr. 2019 à 23:02, Mikhail Naletov ***@***.***> a écrit : Hello! Can I somehow skip the page? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[dcramer]

This is resolved in sentry master and will be fixed in 9.1.1 which we're hoping to get out later today.

See also #12778 and getsentry/self-hosted#167

[basitsattar]

I am using onpremise repo and still getting the same error.

[dcramer]

Are you using 9.1.1?

[ahalf-yuan]

I change to 9.0.0, and it's ok.

[DaveIW2034]

I am using onpremise repo and still getting the same error.
change to 9.0.0

[uhlhosting]

Same issue here

[Fogapod]

I get same error saving settings on admin panel. I'm running On-Premise docker 9.1.2.
Error: AssertionError u'system.url-prefix' cannot be changed at runtime because it is configured on disk

[salty-horse]

I also have the system.url-prefix error with 9.1.2. I changed it in config.yml, restarted sentry, and when I try to change an unrelated setting in the /manage/settings/ screen, it returns the exception. (The url-prefix change did apply itself correctly, and is used in emails sentry sends.)

[BYK]

@salty-horse if you can replicate this in Sentry 10, can you file a new issue please?

[salty-horse]

The issue was that I tried setting system.url-prefix in config.yml to a value different that what was in the database's system.url-prefix option. When I opened the settings screen, the config.yml value was populated in the "Root URL" field, and that field was disabled. Pressing "Save changes" apparently tries to save that new value.

Maybe I wasn't supposed to do that, but nothing in the documentation, the live error message ("error saving settings" or somesuch), or the sentry exception hinted at what was wrong.

[BYK]

@salty-horse I think that behavior is definitely a bug as a disabled field should not affect saving other fields. I'm just curious about whether this still exists in v10 or not, that's why I asked.

[salty-horse]

Sorry, I'm not in a position to test Sentry 10 right now :( Hopefully I described the problem in enough detail in case others encounter the same issue, and don't know how to solve it, and for any willing testers to check if it still occurs.

[NaveenGovindraj]

Hello All,

I have just installed Sentry On-premise 10 (sentry, version 10.1.0.dev0) and I also face the same error that @Fogapod and @salty-horse seem to be facing.

My Scenario:
I wrongly set the system.url-prefix (https://sentry.organization.com) while configuring Sentry on our on-premise server, and I was looking forward to changing it. When I tried to execute the below command(example):
sudo docker-compose run --rm web sentry config set system.url-prefix http://sentry.neworganization.com

I get a similar error message stating:
AssertionError: u'system.url-prefix' cannot be changed at runtime because it is configured on disk.

If my scenario is not relevant to this topic in discussion, kindly advise me on how to resolve this issue.
Please let me know if any further information is needed from my end
Thanks in advance.

Best Regards
Naveen

[BYK]

@NaveenGovindraj it is not related. You need to set this value in your config files and rebuild the sentry container image via docker-compose build