HTTP Error 500 + "Could not load roles" #558
Comments
|
Can you please check if you see a line like the following on the
|
|
Yes, I do! |
|
And after a few minutes doing nothing, these show up in the settings pod: |
|
In my case I reset the data on the storagesystem-data PVC. If you do that, you'd also need to reset the storagesystem cache. If you use an external NATS installation (see ocis-charts/deployments/ocis-nats/helmfile.yaml Lines 6 to 10 in 9023c40 kubectl exec -n ocis-nats deployments/nats-box -- nats stream prune --all KV_storage-system
|
|
Using the deployment example with external When I look at the directory, it is indeed empty: ~ $ ls -la /var/lib/ocis/storage/users/indexes/by-user-id/
total 8
drwx------ 2 ocis-use ocis-gro 4096 May 16 13:32 .
drwx------ 5 ocis-use ocis-gro 4096 May 16 13:32 ..
~ $Inside
What do you mean by that? I try to deploy ocis from scratch, so I destroy everything and deploy again with some changed values. I shouldn't need to reset. Or do I? Also when I use ~ $ curl --verbose http://minio.minio-hot.svc.cluster.local
* Host minio.minio-hot.svc.cluster.local:80 was resolved.
* IPv6: (none)
* IPv4: 10.96.121.235
* Trying 10.96.121.235:80...
* Connected to minio.minio-hot.svc.cluster.local (10.96.121.235) port 80
> GET / HTTP/1.1
> Host: minio.minio-hot.svc.cluster.local
> User-Agent: curl/8.5.0
> Accept: */*
>
< HTTP/1.1 403 Forbidden
< Accept-Ranges: bytes
< Content-Length: 254
< Content-Type: application/xml
< Server: MinIO
< Strict-Transport-Security: max-age=31536000; includeSubDomains
< Vary: Origin
< Vary: Accept-Encoding
< X-Amz-Id-2: 0a0da9bb6b6d3c38dfeab2cc7e8e6ed0fc8ccf27dcf29c14fd96819b66189b4d
< X-Amz-Request-Id: 17CFFCA1BA8B8480
< X-Content-Type-Options: nosniff
< X-Xss-Protection: 1; mode=block
< Date: Thu, 16 May 2024 13:55:54 GMT
<
<?xml version="1.0" encoding="UTF-8"?>
* Connection #0 to host minio.minio-hot.svc.cluster.local left intact
<Error><Code>AccessDenied</Code><Message>Access Denied.</Message><Resource>/</Resource><RequestId>17CFFCA1BA8B8480</RequestId><HostId>0a0da9bb6b6d3c38dfeab2cc7e8e6ed0fc8ccf27dc
~ $The curl above shows up in the Minio Trace, but nothing else from ocis is showing up. Hmmm! |
|
When I combine the |
|
Switching from the builtin nats to the external nats removes all state (thus acting like the purge command I provided).
You shouldn't. But if you assume that you're deploying from scratch, you should challenge / verify if that's really the case. Eg. you should make sure that PersistentVolumes are gone and in the case you use Minikube with the Hostpath provisioner, you actually need to clean up manually ( |
|
I think I found the root cause for these issues in my deployment. I run Longhorn and their Thanks @wkloucek for your assistance! |
I could need some pointers on how to further debug this. I try to deploy this chart, but am stuck at "Internal Server Error". It seems to be connected to when
storageusersis set tos3ng. If I set it toocisand redeploy the chart from scratch, I can login.First some logs and then my config:
ingress-nginx
proxy-..-pod:
{"level":"info","service":"proxy","proto":"HTTP/1.1","request-id":"00380f3e-52b6-4dc2-a8db-7d5cb2ab88da","traceid":"503282f7958ea7331a359732389d547e","remote-addr":"12.34.56.78","method":"GET","status":200,"path":"/ocs/v1.php/cloud/capabilities","duration":3.193941,"bytes":3105,"time":"2024-05-13T21:16:18Z","line":"github.com/owncloud/ocis/v2/services/proxy/pkg/middleware/accesslog.go:34","message":"access-log"} {"level":"info","service":"proxy","proto":"HTTP/1.1","request-id":"f9c20b29dd32806513a4b7693748717e","traceid":"67c9e2f9d84664033b44c34b48e7fcb5","remote-addr":"xxx.xxx.63.230","method":"GET","status":200,"path":"/konnect/v1/userinfo","duration":37.805854,"bytes":265,"time":"2024-05-13T21:16:18Z","line":"github.com/owncloud/ocis/v2/services/proxy/pkg/middleware/accesslog.go:34","message":"access-log"} {"level":"error","service":"proxy","error":"{\"id\":\"com.owncloud.api.settings\",\"code\":502,\"detail\":\"circuit breaker is open\",\"status\":\"Bad Gateway\"}","time":"2024-05-13T21:16:18Z","line":"github.com/owncloud/ocis/v2/services/proxy/pkg/userroles/defaultrole.go:38","message":"Could not load roles"} {"level":"error","service":"proxy","error":"{\"id\":\"com.owncloud.api.settings\",\"code\":502,\"detail\":\"circuit breaker is open\",\"status\":\"Bad Gateway\"}","time":"2024-05-13T21:16:18Z","line":"github.com/owncloud/ocis/v2/services/proxy/pkg/middleware/account_resolver.go:140","message":"Could not get user roles"} {"level":"info","service":"proxy","proto":"HTTP/1.1","request-id":"6310a8cf-ecac-41ab-94bf-fb6e65cf7189","traceid":"0a15fd98788bd0f25562bf66bd5b62dd","remote-addr":"12.34.56.78","method":"GET","status":500,"path":"/ocs/v1.php/cloud/user","duration":70.088662,"bytes":0,"time":"2024-05-13T21:16:18Z","line":"github.com/owncloud/ocis/v2/services/proxy/pkg/middleware/accesslog.go:34","message":"access-log"} {"level":"info","service":"proxy","proto":"HTTP/1.1","request-id":"64f0ace2a798d5b14ec4bfd177c2af50","traceid":"f3ba7e9868dedf663bcb42f7b76267f9","remote-addr":"12.34.56.78","method":"GET","status":200,"path":"/themes/owncloud/assets/logo.svg","duration":0.923335,"bytes":9896,"time":"2024-05-13T21:16:18Z","line":"github.com/owncloud/ocis/v2/services/proxy/pkg/middleware/accesslog.go:34","message":"access-log"}Here it looks like an Gateway Error (502) which results in "Could not load roles".
My
values.yaml:values.yaml
There are quiet a few annotations. That's because I first got HTTP Error
431, too large header. Managed to get past that. But now I am at 500 or 502. As mentioned, probably connected tostorageusersdriver setting.One thing I wounder if there exists a
insecuresetting for the s3ng driver? Maybe the 502 error comes from ocis not being able to verify the minio certificate?Any hints?
The text was updated successfully, but these errors were encountered: