Skip to content
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

db_bench: align the tests we're using #325

Closed
Yuval-Ariel opened this issue Jan 4, 2023 · 9 comments · Fixed by #355
Closed

db_bench: align the tests we're using #325

Yuval-Ariel opened this issue Jan 4, 2023 · 9 comments · Fixed by #355
Assignees
Labels
bug Something isn't working Working on it A feature request that is being actively worked on

Comments

@Yuval-Ariel
Copy link
Contributor

Describe the bug
when used with multiple cfs, the test will only use the default cf.

Additional context
this has been addressed before but must've slipped during rebasing

@Yuval-Ariel Yuval-Ariel added the bug Something isn't working label Jan 4, 2023
@Yuval-Ariel Yuval-Ariel self-assigned this Jan 4, 2023
@Yuval-Ariel
Copy link
Contributor Author

Yuval-Ariel commented Jan 6, 2023

also fix BGWriter, readrandom, seekrandom and readrandomwriterandom.
made sure these tests:

  1. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
  2. report the bytes throughput
  3. ability to have a read and write limiter where ever relevant.
  4. operations are done on all cfs randomly and not only on the default cf

@Yuval-Ariel Yuval-Ariel linked a pull request Jan 6, 2023 that will close this issue
Yuval-Ariel added a commit that referenced this issue Jan 10, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
@Yuval-Ariel Yuval-Ariel changed the title db_bench: seekrandomwriterandom is only using default cf db_bench: align the tests we're using Jan 15, 2023
@Yuval-Ariel
Copy link
Contributor Author

@erez-speedb , plz run performance. we're expecting no change in all tests. the only change that should be visible is in the seekrandomwriterandom which will use all the cfs in huge db test and the seekradomwhilewriting now writes to all cfs

@Yuval-Ariel Yuval-Ariel added the Working on it A feature request that is being actively worked on label Jan 23, 2023
@Yuval-Ariel
Copy link
Contributor Author

@erez-speedb , status?

@erez-speedb
Copy link

@Yuval-Ariel The multi CF is running but I'm not sure what performance is there to test here?

@Yuval-Ariel
Copy link
Contributor Author

its a matter of correctness.. but if theres a performance difference in the single cf tests then its bad

@erez-speedb
Copy link

Multi CF test,
The difference is only with the seekrandomwriterandom tests
https://admin.speedb.io/performance?items=nl1nwB81l5DDRhuzwWjd&items=hKWqp0Aknv7eR3CDlnnG&colors=%23F06292&colors=%2300796b

Image

@erez-speedb
Copy link

The small DB showed degradation with the seekrandomwriterandom tests.
Running the test and the large obj test again

Image

@erez-speedb
Copy link

@Yuval-Ariel I tested few times and both large obj and small DB tests shows the degradation
https://admin.speedb.io/performance?items=IxsAvKvunCafENia4udE&items=DQbr0ntUjYJDz3RLExzO&colors=%23F06292&colors=%23BA68C8
And
https://admin.speedb.io/performance?items=IxsAvKvunCafENia4udE&items=DQbr0ntUjYJDz3RLExzO&colors=%23F06292&colors=%23BA68C8

This is unless the degradation is from introducing the rate limit, and then it is part of a separate discussion.
For that, I'm testing the old main again.

@erez-speedb
Copy link

The above degradation is from the rate limit

Image

Past performance tests.

Yuval-Ariel added a commit that referenced this issue Feb 6, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
ayulas pushed a commit that referenced this issue Feb 26, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
Yuval-Ariel added a commit that referenced this issue May 1, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
Yuval-Ariel added a commit that referenced this issue May 4, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
udi-speedb pushed a commit that referenced this issue Nov 13, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
udi-speedb pushed a commit that referenced this issue Nov 15, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
udi-speedb pushed a commit that referenced this issue Dec 4, 2023
make sure all the tests used:
1. are working on all cfs if theres more than 1
2. use one rand num to get the db and cf and the key which makes it possible to find a key in a cf later on.
3. have the ability for read and write limiter wherever relevant.

tests:
seekrandomwriterandom, BGWriter, readrandom, seekrandom and readrandomwriterandom
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working Working on it A feature request that is being actively worked on
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants