You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Our cluster_dump functionality is currently broken wrt. key roundtrips (#8540), which we need to fix. Since I do not see the added value in supporting both msgpackandyaml, I'm inclined to limit the fix for roundtripping to msgpack and drop support for yamlwithout a deprecation cycle. (It's mostly broken anyways.) Note that we will still have DumpArtefact.to_yamls in case users want to create human-readable files. Also, msgpack is already a required dependency.
The text was updated successfully, but these errors were encountered:
Actually, msgpack was added only later because the dumps grew too large. the benefit of yaml is that it is human readable. I often analyzed those dumps by grepping/searching and reading the files.
I'm fine with dropping support for the dump to be written directly to yaml but there is also DumpArtefact.to_yamls which probably has the same problem
I'm fine with dropping support for the dump to be written directly to yaml but there is also DumpArtefact.to_yamls which probably has the same problem
It will. Since we won't have to be able to roundtrip, I think we can get away with a simpler implementation here, i.e., simply stringifying keys. You could then grep for the stringified key in the yaml files.
Our
cluster_dump
functionality is currently broken wrt. key roundtrips (#8540), which we need to fix. Since I do not see the added value in supporting bothmsgpack
andyaml
, I'm inclined to limit the fix for roundtripping tomsgpack
and drop support foryaml
without a deprecation cycle. (It's mostly broken anyways.) Note that we will still haveDumpArtefact.to_yamls
in case users want to create human-readable files. Also,msgpack
is already a required dependency.The text was updated successfully, but these errors were encountered: