feat(shed): lotus-shed sectors dump-sectors #12703
Conversation
rvagg
force-pushed
the
rvagg/dump-sectors
branch
from
3a9980f
to
c9a427d
Compare
| } | ||
|
|
||
| path, err := lkrepo.SplitstorePath() |
There was a problem hiding this comment.
What is going on here ? What was breaking that made you change this?
There was a problem hiding this comment.
When you open universal blockstore it should be wrapped around all of the splitstore right?
There was a problem hiding this comment.
@ZenGround0: when using the original I couldn't get it to run - when it initialised and figured out what the latest head is (from the metadata store) it would then try and load that tipset but fail with not-found errors. Switching to this makes it work. My assumption is that this works fine for a non-splitstore setup, but breaks when using a splitstore, and nobody running splitstore has been using the commands that use this code until I tried it.
Both patterns are found throughout lotus-shed, I copied a lot of this new stuff from elsewhere. I'd like to understand it a bit better to make it consistent throughout, maybe just make a helper function for this. But I haven't taken the time to bother going that deep yet. If you have more of a clue then I'm all ears!
Most of the fields in SectorOnChainInfo for all of the records on chain (using the power actor to list miners and then listing all of their sectors). No filtering so far, output as CSV. Currently doesn't print the two CIDs. This came up as part of the FIP-0096 (or whatever number this will get filecoin-project/FIPs#1079) discussions looking at termination fees and I wanted to collect some data for it.