stanchion job of riak is crashing about cf-riak-cs-release HOT 7 CLOSED

We have created an issue in Pivotal Tracker to manage this. You can view the current status of your issue at: https://www.pivotaltracker.com/story/show/82739708.

from cf-riak-cs-release.

Is this response from curling riak-cs directly (which I would expect to return a 403 as authentication with S3 API is complex), or from the broker logs?

from cf-riak-cs-release.

This response is from the broker logs and it is the same by curl the broker. I am confused because of getting the catalog is working and doesn't give a 403. If there is an authentication issue then I would expect to get 403 for the catalog as well.

from cf-riak-cs-release.

I don't believe that authenticating with the broker is the problem. "InvalidAccessKeyId" indicates the key id and secret are not correct for the broker to talk to the backend.

Please confirm in your manifest that the key and secret configured for the broker job matches the key and secret for the riak-cs job and the stanchion jobs.

from cf-riak-cs-release.

I am going to check it, but my assumption is that 403 is only a misleading follow up error. In meantime I found out that a Riak node seems not to have enough memory which causes problem with stanchion. I am trying a different deployment and will see...

from cf-riak-cs-release.

Looking at this issue (a bit delayed) and I'm not sure we have enough information to reproduce. Did you get this resolved on your own?

If not, do you have the info from googling that lead you to believe it was stanchion?
We'll probably also need a sanitized version of your deployment manifest if we're going to debug this.
You might also try v7, but I don't know of any stanchion related changes recently.

from cf-riak-cs-release.

Closing due to inactivity.
Feel free to create a new ticket if you have new information.

from cf-riak-cs-release.