|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
frostyx commented 4 years ago | ||
praiskup commented 4 years ago Wouldn't it be propagated up then, and than it would be up to the caller to take care of that? I don't have problem to add except, log something, and re-raise; but it seems to me that we track those exceptions in | ||
frostyx commented 4 years ago You are right, I didn't realize that. Scratch that idea. | ||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
This is to prevent "no space left on device" on / partition.
Since we recycle user-dedicated builder machines heavily, we need
to clean directories like
$ ls -1 /var/cache/mock BUILD_ID-fedora-rawhide-x86_64 BUILD_ID-fedora-rawhide-x86_64-bootstrap
which contain yum/dnf caches there (soon to be enabled), and
mock's root_cache.
If we don't do this, after several (tens?) of builds on the same
builder, the / partition might be wasted. This is especially
important in Fedora Copr on aarch64 where we have small disks.
You are doing try-finally, so you probably expect, that the rpm producing code can fail. What do you think about adding an
except
block and at least logging the exception?