On 07/13/2017 06:30 AM, Chris Murphy wrote:
On Wed, Jul 12, 2017 at 10:22 PM, Samuel Sieb <samuel@xxxxxxxx> wrote:
On 07/12/2017 09:14 PM, Chris Murphy wrote:
On Wed, Jul 12, 2017 at 10:03 PM, Chris Murphy <lists@xxxxxxxxxxxxxxxxx>
wrote:
-rw-rw-r--. 1 chris chris unconfined_u:object_r:user_home_t:s0
53190 Jul 12 19:26 kernel-4.11.10-300.fc26.x86_64.rpm
kernel-4.11.10-300.fc26.x86_64.rpm: data
I downloaded the same files, from the same location on the same
Rawhide system again and now it's
kernel-4.11.10-300.fc26.x86_64.rpm: RPM v3.0 bin
i386/x86_64 kernel-4.11.10-300.fc26
I don't understand why, or why it would behave one way for hours and
now behave a different way.
It would be interesting to see what the other contents of the file were, but
I guess it's too late now. Are they the same size as before?
HUH. It gets even stranger! The ones first downloaded that are broken,
every single one is exactly 8192 bytes smaller than the good
replacements.
Do you happen to run F26?
The phenomenon you are describing would match with what I am
occasionally observing happening on F26.
What I am occasionally seeing here, is writing out the "end of a (large)
file (last block?) to disk" to take a very long time (In the order of
several minutes), seemingly occasionally leaving shortened files behind.
To me, the symptoms look like fflush/fclose/fsync issues which lets me
suspect the origin being the kernel or glibc.
Ralf
_______________________________________________
devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
