-
Notifications
You must be signed in to change notification settings - Fork 394
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
docs: locking-related updates #868
Conversation
dvc run/repro
dvc run/repro
dvc run/repro
dvc run/repro
dvc run/repro
dvc run/repro
@efiop still WIP? |
Yes, I am a bit ashamed of this PR, so made it WIP to take another look 🙂Will give it another look till tomorrow and will address your concerns. |
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.
In addition to previous comments, it would be great if we can link existing instanced of "always locked" phrase in the few docs where these exist. Thanks
…oject (instead of downloading) per iterative#859 (review), + remove "s around 'locked` in some docs (related to iterative#860 and iterative#868).
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
28dc9a3
to
5d007cc
Compare
dvc run/repro
dvc run/repro
@jorgeorpinel Not sure what you mean. Could you clarify? |
Updated this PR a bit. Don't want to make it turn into a long one, so I suppose it is better to stay with the current amount of info to start with. Open to suggestions to improve it and make it mergeable. |
My main general concern here at this point is that basic locking is not explained anywhere, I think. So talking about how locking still allows parallel execution seems like skipping a step 🙂 Please see #868 (review) |
@jorgeorpinel I think it's a good first step for the "document multiple parallel repros/runs" - that's what this PR is essentially about, not "document locking in dvc run/repro" in general the @efiop named it initially. I think we do describe project lock only in the DVC files and directories? It's not enough, true. But it should be probably a separate PR to write a proper DVC internals section aboutt
|
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.
Think about making the section name more explicit.
Some clarification in the dvc repro
doc can be beneficial.
Sorry guys, will get back to this later(hopefully tomorrow). It is in my saved tickets. |
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 agree with Ivan's review so approving on my end.
I think we should prioritize #876 as well though – cc @shcheklein
3861fbb
to
1b1c1b2
Compare
dvc run/repro
dvc run/repro
Sorry for the delay. I've updated the docs to separate "running other commands parallel" from "parallel stage execution". Should be more clear now. |
@@ -52,6 +52,14 @@ captures data and <abbr>caches</abbr> relevant <abbr>data artifacts</abbr> along | |||
the way. See [this example](/doc/get-started/example-pipeline) to learn more and | |||
try creating a pipeline. | |||
|
|||
### Running other dvc commands in parallel |
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'm not sure about this ... why do we mention this here and in repro and not in dvc add
?
We also def missing any reasonable context about the project lock
.
It feels that it can be a general user guide
section that describes locking and when and how do we release it. How to find those files, etc.
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.
@shcheklein because dvc add
doesn't release repo lock, but repro/run
do release it. So for add
the section would be running dvc run/repro in parallel
which is odd to say the least. Eventually we will release repo lock on checksum computations/copies too, so we will alter the docs accordingly.
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.
Ok, I guess this behaviour is simply an expected one, hence why so many people were running into it. So it is not worth documenting it here. Will move to user-guide.
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.
Ok, removed that note entirely. Will wait until we do similar things on hash computation/copying/downloading and will then create a locking
article, that would be too incomplete to bother with right now.
See | ||
[Running other dvc commands in parallel](/doc/command-reference/run#running-other-dvc-commands-in-parallel). | ||
|
||
### Parallel stage execution |
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.
this is a good one! I'll @jorgeorpinel review specific language nuances, but I hope it will be useful for people. (When we at last have a How to section it can be probably moved there with a link to it from this document - just to keep a sane size)
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.
How to section?
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.
"How to" (name can change) - we have a bunch of tickets already that start like "write a How to **** "
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 see. OK, I created #899 to determine this.
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.
parallel stage execution
is great!
locking looks strange, w/o context - I doubt users would get it - check my comment with some suggestions on how to move forward with it.
Removed locking explanation entirely for now, will wait until we implement the same thing for hash/copying/downloading and will then create an article about it in the user-guide. If I try to do it now it will be too incomplete. Now this PR introduces a note about parallel execution to |
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.
Rephrasing of text in Parallel stage execution:
I had this same impression before (see #868 (review)). Do you want to assign a priority to #876 so we address it sooner than later @shcheklein?
@efiop you can probably just reuse #876 for this. Feel free to update that issue's description or add a comment. |
Just one last comment from me, I think ^ |
@jorgeorpinel please, feel free to merge whenever you feel it's ready. @efiop thanks!! |
Related to iterative#860
Related to #860
Disregard the recommendations below if you use Edit on GitHub button to improve the docs in place.
❗ Please read the guidelines in the Contributing to the Documentation list if you make any substantial changes to the documentation or JS engine.
🐛 Please make sure to mention
Fix #issue
(if applicable) in the description of the PR. This enables GitHub to link the PR to the corresponding bug and close it automatically when PR is merged.Thank you for the contribution - we'll try to review and merge it as soon as possible. 🙏