On 7/1/20 11:16 AM, Rachel Sibley wrote: > > > On 7/1/20 12:42 PM, Jens Axboe wrote: >> On 7/1/20 10:37 AM, Rachel Sibley wrote: >>> Hi, we're seeing multiple panics across all arches, I included a snippet of the call trace for both >>> xfstests and boot test. >>> >>> You should be able to inspect in more detail by viewing the console.log under each build/tests directory: >>> https://cki-artifacts.s3.us-east-2.amazonaws.com/index.html?prefix=datawarehouse/2020/06/30/609250 >> >> This was due to a bad patch series, which since got reverted and redone. Current >> tree should be fine. >> >> Now it doesn't matter for this one since I guessed what this was and found it >> before the bot did, but I do wish the reports were easier to look at. I should >> not have to dig through directories (which were empty when the report went out, > > Sorry about that we noticed this right after we sent the report and > worked quickly to resolve it on our end, the logs are now accessible > in the external artifacts location. I was probably just too quick, but if we can fix the below, then it'd work much nicer and the logs would just be a secondary resource. >> btw) to find logs, then download logs and leaf through hundreds of kb of text >> to find out why the bot thought the tree was broken. It should be readily >> apparent and in the email. If there's an OOPS, include the oops. > > Agreed, this is also something we'd like to do and we have an > outstanding ticket to work on it. I'll follow up and see if we can > move this along quicker to make it easier to find it in the reports. Thanks, that would be a massive improvement! The OOPS is really the key thing here, and then I think it's fine to have to dig in logs/directories to find other related information. Sometimes you just know what it is just by seeing the OOPS. -- Jens Axboe
