I made a new job in our jenkins to build dizzy; in first build it got:

framebuffer fsl-image-machine-test@imx6slevk (10/19) ERROR: Unable to install packages. Command '.../build-framebuffer/tmp/sysroots/x86_64-linux/usr/bin/smart --quiet --data-dir=.../build-framebuffer/tmp/work/imx6slevk-poky-linux-gnueabi/fsl-image-machine-test/1.0-r0/rootfs/var/lib/smart install -y packagegroup-fsl-tools-gpu@imx6slevk packagegroup-core-tools-testapps@imx6slevk packagegroup-base-extended@imx6slevk run-postinsts@all packagegroup-fsl-tools-benchmark@imx6slevk packagegroup-core-tools-profile@imx6slevk packagegroup-fsl-gstreamer-full@imx6slevk packagegroup-fsl-tools-gpu-external@all packagegroup-fsl-gstreamer@imx6slevk packagegroup-fsl-tools-testapps@imx6slevk packagegroup-core-boot@imx6slevk locale-base-en-us@cortexa9hf_vfp_neon locale-base-en-gb@cortexa9hf_vfp_neon' returned 1:
framebuffer fsl-image-machine-test@imx6slevk (10/19) error: Failed to download packages:
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/lib-mpeg2-dec-arm11-elinux2-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Invalid SHA (expected 2adb6bb5ed06a3a9c16495812e7b5c0c811398af, got 182ffdc3af75e20b1edef3f9a017567060a79060)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/all/packagegroup-fsl-tools-gpu-external-1.0-r0.all.rpm: Invalid SHA (expected b4a4ea55cf4874f6e6ae4ce9865564671c29f395, got 38da9f444d659b41a43502ab68773714cde7f8bb)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/all/ca-certificates-20140325-r0.all.rpm: Unexpected size (expected 182137, got 182129)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/lib-peq-arm11-elinux1.fhw-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Invalid SHA (expected 870ce989ddd67ad9af2f9a54883a7398cb946b8f, got 1d1a417fb98cf041c38b776e15875cde7551fe73)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/lib-oggvorbis-dec-arm11-elinux2-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 106423, got 106422)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/libfslcodec-wb-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 153001, got 153004)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/libfslcodec-aac-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 154423, got 154425)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/libfslcodec-nb-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 335260, got 335263)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/libfslcodec-h264-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 145115, got 145117)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/libfslcodec-mp3-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 497617, got 497611)
framebuffer fsl-image-machine-test@imx6slevk (10/19) error:     .../build-framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/libfslcodec-mpeg4asp-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Unexpected size (expected 226341, got 226342)

You have a damaged rpmdb (is all that can be told from the reported details).

Running out of diskspace is one common cause of truncated headers,
that is consistent with (failed to download messages unrelated to
rpm itself).

disabling fsync(2) (for performance) and exceptional evens like
reboots or turning off VM's is another cause of truncated I/O.

I'm not sure how/why I own this bug ... I need a more complete
log than given on why the database was already damaged when the
jenkins build was started.

Otavio, can you update us on this, are you still seeing this issue?  If not can we close it?

framebuffer/tmp/deploy/rpm/cortexa9hf_vfp_neon_mx6/lib-peq-arm11-elinux1.fhw-4.0.1-r0.cortexa9hf_vfp_neon_mx6.rpm: Invalid SHA (expected 870ce989ddd67ad9af2f9a54883a7398cb946b8f, got 1d1a417fb98cf041c38b776e15875cde7551fe73)

I get, from time to time for both rpm and opkg backends.

that message I think is coming from RPM directly..

The only thing I can think of is that you have a race condition and that thing is being rebuilt (replaced?) which the indexing or other install components are running.  Causing the initial hash to be read, and then the file is updated and a subsequent is read that do not match.

Yes; I have the very same theory however this is not easy to reproduce.

That isn't an RPM message (though is formatted similarly). The digest would be named more specifically than "SHA" if from RPM.

Saul,

It appears the answer has been given.  You can put this back into NEEDINFO if you have another question.

Talked with Paul, assign to me atm.

Does anyone see this recently, please ?

(In reply to comment #11)
> Does anyone see this recently, please ?

I did not see this for a while.

Smart is gone, now dnf.