-
Notifications
You must be signed in to change notification settings - Fork 3.8k
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
kv: remove clock update on BatchResponse #76233
kv: remove clock update on BatchResponse #76233
Conversation
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.
Stared at the comment a bunch but still confused. Attempt at a re-wording to highlight my particular confusions:
- The first is that a BatchRequest can legitimately remain ahead of the local HLC clock if it contains a synthetic timestamp ("local timestamp") tomorrow. For example, <actually give an example, I'm too confused right now, what do you mean by this bit about "as long as any clock reading"..?>
- The second is that even for requests without such timestamps, the response timestamp would only ever differ from the incoming timestamp by one logical tick, so... (and here I'm confused again, if you say "they don't" then why is that ok, is this synthetic bit again?)
I'm admittedly not anywhere near my brightest moment right now, but I'd appreciate a comment that caters to me in my current state as well 🥲
Reviewable status: complete! 0 of 0 LGTMs obtained
Before this change, we were updating the local clock with each BatchResponse's WriteTimestamp. This was meant to handle cases where the batch request timestamp was forwarded during evaluation. This was unnecessary for two reasons. The first is that a BatchResponse can legitimately carry an operation timestamp that leads the local HLC clock on the leaseholder that evaluated the request. This has been true since cockroachdb#80706, which introduced the concept of a "local timestamp". This allowed us to remove the (broken) attempt at ensuring that the HLC on a leaseholder always leads the MVCC timestamp of all values in the leaseholder's keyspace (see the update to `pkg/kv/kvserver/uncertainty/doc.go` in that PR). The second was that it was not even correct. The idea behind bumping the HLC on the response path was to ensure that if a batch request was forwarded to a newer timestamp during evaluation and then completed a write, that forwarded timestamp would be reflected in the leaseholder's HLC. However, this ignored the fact that any forwarded timestamp must have either come from an existing value in the range or from the leaseholder's clock. So if those didn't lead the clock, the derived timestamp wouldn't either. It also ignored the fact that the clock bump here was too late (post-latch release) and if it had actually been needed (it wasn't), it wouldn't have even ensured that the timestamp on any lease transfer led the maximum time of any response served by the outgoing leaseholder. There are no mixed-version migration concerns of this change, because cockroachdb#80706 ensured that any future-time operation will still continue to use the synthetic bit until all nodes are running v22.2 or later.
b9ee582
to
35151c9
Compare
@tbg thanks for taking a look and providing suggestions. I've rewritten the commit message to more clearly explain why this is unnecessary now and also was broken before when we thought it was necessary. PTAL. |
bors r+ |
Build succeeded: |
Before this change, we were updating the local clock with each BatchResponse's WriteTimestamp. This was meant to handle cases where the batch request timestamp was forwarded during evaluation.
This was unnecessary for two reasons.
The first is that a BatchResponse can legitimately carry an operation timestamp that leads the local HLC clock on the leaseholder that evaluated the request. This has been true since #80706, which introduced the concept of a "local timestamp". This allowed us to remove the (broken) attempt at ensuring that the HLC on a leaseholder always leads the MVCC timestamp of all values in the leaseholder's keyspace (see the update to
pkg/kv/kvserver/uncertainty/doc.go
in that PR).The second was that it was not even correct. The idea behind bumping the HLC on the response path was to ensure that if a batch request was forwarded to a newer timestamp during evaluation and then completed a write, that forwarded timestamp would be reflected in the leaseholder's HLC. However, this ignored the fact that any forwarded timestamp must have either come from an existing value in the range or from the leaseholder's clock. So if those didn't lead the clock, the derived timestamp wouldn't either. It also ignored the fact that the clock bump here was too late (post-latch release) and if it had actually been needed (it wasn't), it wouldn't have even ensured that the timestamp on any lease transfer led the maximum time of any response served by the outgoing leaseholder.
There are no mixed-version migration concerns of this change, because #80706 ensured that any future-time operation will still continue to use the synthetic bit until all nodes are running v22.2 or later.