You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is related with the issue #11111.
We had just faced that issue in our kubernetes environment (we don't use helm charts there) and, as it was said in the previous issue, the root cause was related to minIO configs that were corrupted.
I want to share the way to fix it, as in the other issue some people suggest to recreate the volume so the configs were created again, but that was not possible in our env, and I guess is not always posible to delete the existing volume.
We tried to manually set the cors config again using the minIO CLI, and there is how the solution to the problem appeared!
user@user-mbp minio % mc admin config set dc-fs api "cors_allow_origin=*"
mc: <ERROR> Unable to set 'api cors_allow_origin=*' to server: found invalid keys (extend_list_cache_life=0s ) for 'api' sub-system, use 'mc admin config reset myminio api' to fix invalid keys.
user@user-mbp minio % mc admin config reset dc-fs api
Key is successfully reset.
user@user-mbp minio % mc admin config set dc-fs api "cors_allow_origin=*"
Successfully applied new settings.
How we narrow down the issue to a minIO config issue that only appears after the upgrade?
CORS wasn't working in one env that was updated some time ago, but it was working fine in another env that was still using a previous minIO version. We upgrade that env that was working fine, and we started facing the CORS issue there, meaning that after the upgrade, minIO brakes the CORS config in some way. But, using that new minIO version with a fresh installation, that was working fine. So the problem only raised after the upgrade.
mc: Unable to set 'api cors_allow_origin=*' to server: found invalid keys (extend_list_cache_life=0s ) for 'api' sub-system, use 'mc admin config reset myminio api' to fix invalid keys.
Let me check what challenges are here and if we can automatically handle it.
This is related with the issue #11111.
We had just faced that issue in our kubernetes environment (we don't use helm charts there) and, as it was said in the previous issue, the root cause was related to minIO configs that were corrupted.
I want to share the way to fix it, as in the other issue some people suggest to recreate the volume so the configs were created again, but that was not possible in our env, and I guess is not always posible to delete the existing volume.
We tried to manually set the cors config again using the minIO CLI, and there is how the solution to the problem appeared!
How we narrow down the issue to a minIO config issue that only appears after the upgrade?
CORS wasn't working in one env that was updated some time ago, but it was working fine in another env that was still using a previous minIO version. We upgrade that env that was working fine, and we started facing the CORS issue there, meaning that after the upgrade, minIO brakes the CORS config in some way. But, using that new minIO version with a fresh installation, that was working fine. So the problem only raised after the upgrade.
Originally posted by @harshavardhana in #11111 (comment)
The text was updated successfully, but these errors were encountered: