12127 2017-09-26 13:47:26 +0000 Duplicate log observed on oe-selftest runs 2017-11-09 13:27:22 +0000 1 1 10 QA/Testing Build Testing general 2.4 x86 Multiple RESOLVED FIXED Medium+ normal 2.5 M1 1 leonardo.sandoval.gonzalez jair.de.jesus.gonzalez.plascencia anibal randy.macleod richard.purdie ross.burton --- --- No (bug/feature does not impact docs) oldest_to_newest 77139 0 leonardo.sandoval.gonzalez 2017-09-26 13:47:26 +0000 In recent runs, oe-selftest log have duplicate lines as observed in [1]: https://autobuilder.yocto.io/builders/nightly-oe-selftest/builds/546/steps/Running%20oe-selftest/logs/stdio 77140 1 richard.purdie 2017-09-26 14:18:12 +0000 They seem to come from runqemu and for each runqemu run, a new level of duplication is added. This happens sometimes and not others, generally on newer autobuilders but doesn't always happen. 77737 2 ross.burton 2017-10-20 14:37:19 +0000 *** Bug 12257 has been marked as a duplicate of this bug. *** 78034 3 richard.purdie 2017-11-07 17:50:55 +0000 Starting to suspect this happens when there is a server reconnection. Each reconnection adds extra log duplication. 78035 4 anibal 2017-11-07 18:00:26 +0000 (In reply to comment #3) > Starting to suspect this happens when there is a server reconnection. Each > reconnection adds extra log duplication. The qemu runner adds a new stream handler [1] when is lunched, the testimage class uses the main logger [2] and the oe-selftest script also initialize its own logger [3]. For me the qemu runner needs to use the logger from outside (testimage, selftest), instead of add every time an StreamHandler [3] that cuases the duplication. [4] [1] http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/meta/lib/oeqa/utils/qemurunner.py#n60 [2] http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/meta/classes/testimage.bbclass#n160 [3] http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/scripts/oe-selftest#n46 [4] http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/meta/lib/oeqa/utils/commands.py#n331 78100 5 richard.purdie 2017-11-09 13:27:22 +0000 http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=a87cae5bb1b0168b29eef7076756ef1d7179ee72 and the preceding 8 commits address this issue. Appears that a python bug was fixed in 3.6 which started exposing these messages. http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=729b1d3afd3d4b8eed4a4f4ddeb27780db47dcd5 in particular.