-
Notifications
You must be signed in to change notification settings - Fork 702
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
Fix cluster info sent stats for message with light header #1563
Conversation
@zuiderkwast let me know if we can get the type without modifying the method signature. |
aefef81
to
54a762f
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## unstable #1563 +/- ##
============================================
+ Coverage 70.94% 70.95% +0.01%
============================================
Files 120 121 +1
Lines 65004 65130 +126
============================================
+ Hits 46116 46215 +99
- Misses 18888 18915 +27
|
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.
Good idea. I replies to your question in a comment below.
We don't need to distinguish the heavy/light variants in the stats counter, because we have some other stats counting the bytes?
I initially thought of adding |
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 initially thought of adding
light_{type}_sent
/light_{type}_received
and keep the messages with regular hdr stats as is. I felt it's not that useful for Valkey users but helps with debugging for us. WDYT ?
I agree it's not useful for users. The light/heavy message type is an implementation detail. The bytes sent/received over the cluster bus can be interesting for users. We have that? I can't see it in the CLUSTER INFO man page.
Shall we mention it as a bugfix in the release notes? Backport to 8.0?
Do you mean the
Yeah, I think we should backport it to 8. |
I wasn't aware of CLUSTER LINK. Thanks. Looks like it shows only currently connected nodes. I guess we could add sent/received bytes there. I was thinking of a total sent/received over the cluster bus, probably in CLUSTER INFO. Perhaps we should wait until anyone actually requests it. :) |
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.
LGTM.
Signed-off-by: Harkrishn Patro <[email protected]>
Signed-off-by: Harkrishn Patro <[email protected]>
Co-authored-by: Binbin <[email protected]> Signed-off-by: Harkrishn Patro <[email protected]> Signed-off-by: Harkrishn Patro <[email protected]>
d5dd58e
to
88d14a3
Compare
…1563) This issue affected only two message types (CLUSTERMSG_TYPE_PUBLISH and CLUSTERMSG_TYPE_PUBLISHSHARD) because they used a light message header, which caused the CLUSTER INFO stats to miss sent/received message information for those types. --------- Signed-off-by: Harkrishn Patro <[email protected]> Signed-off-by: Harkrishn Patro <[email protected]> Co-authored-by: Binbin <[email protected]> Signed-off-by: Harkrishn Patro <[email protected]>
Backport to 8.0 PR: #1571 |
…1563) This issue affected only two message types (CLUSTERMSG_TYPE_PUBLISH and CLUSTERMSG_TYPE_PUBLISHSHARD) because they used a light message header, which caused the CLUSTER INFO stats to miss sent/received message information for those types. --------- Signed-off-by: Harkrishn Patro <[email protected]> Signed-off-by: Harkrishn Patro <[email protected]> Co-authored-by: Binbin <[email protected]> Signed-off-by: proost <[email protected]>
This issue only impacted two message type (CLUSTERMSG_TYPE_PUBLISH / CLUSTERMSG_TYPE_PUBLISHSHARD) as those were the only message type uses light message header where the
CLUSTER INFO
stats had missing information of sent/received messages of that type.Before
After