-
Notifications
You must be signed in to change notification settings - Fork 923
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat(share): EDSStore scaffolding #1232
Conversation
Codecov Report
@@ Coverage Diff @@
## main #1232 +/- ##
==========================================
- Coverage 55.42% 55.12% -0.31%
==========================================
Files 178 180 +2
Lines 10746 10890 +144
==========================================
+ Hits 5956 6003 +47
- Misses 4211 4288 +77
- Partials 579 599 +20
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. |
e40c45e
to
3c45fb1
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
- can you doc all noops in blockstore and why they're noops (even if it's just a quick sentence and say check ADR for more info)
334b6ea
to
c30a3aa
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I deep review of EDStore only. I witsh Blockstore was a separate PR.
General comments for the EDSStore:
- We should wrap errors in each method with meaningful messages via
fmt.Errorf(%w)
- We should start a habit of doing this more often. Our codebase lacks this, and usually, it is hard to grasp what happened from the error message and such wrappings should improve this.
- Each result chan read must obey Context. Check one of the comments for more details on that
share.Root
is actually a full blown DAHeader. The networking layer we build on top, won't have access to it, as we don't want to send it over the wire. Thus, we have to change it to theDataHash
orDAHeader.Hash
result. Initially, I wanted to ask to do the change in this PR, but then decided to do in a subsequent PR after we clarify this in the Storage ADR first. We also need to come up with cool alias name for the DataHash inshare
pkg 😂
14db310
to
29ef5ed
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good, awesome work!
I've added one small optimisation suggestion.
Removed blockstore so it can go into a new PR and can be reviewed more thoroughly, this PR is too big rn and the main part of it can get merged |
4797cbf
to
a8394d3
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There is one general comment still unaddressed from the last review:
- We should wrap errors in each method with meaningful messages via fmt.Errorf(%w)
- We should start a habit of doing this more often. Our codebase lacks this, and usually, it is hard to grasp what happened from the error message and such wrappings should improve this.
E.g. if the Remove method fails, wrappings will hint to us which actions specifically failed, as Remove consists of Destroying the shard, dropping the index, and Removing the OS file. Without wrappings it will unclear which of this actions errored out
Co-authored-by: rene <41963722+renaynay@users.noreply.github.com>
8ce9631
to
48fb084
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Awesome. Thanks man
Co-authored-by: Hlib Kanunnikov <hlibwondertan@gmail.com>
Co-authored-by: Hlib Kanunnikov <hlibwondertan@gmail.com>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good, i've already reviewed this and i guess blockstore has been removed so no real comments other than re topIdx
Based on top of #1158, replaces #1051
Closes #1107
Closes #1108
Closes #1110
Closes #1111
Closes #1112