+1 - thanks for a great write up Matteo, it looks like it'll be a
great improvement!
(Sorry for casting a late vote, but I figure it's worth affirming the
PIP even if we've already closed the vote.)
Thanks,
Michael
On Sun, Jul 24, 2022 at 9:32 PM Matteo Merli wrote:
>
> Closing this vote with 3
Closing this vote with 3 +1s and no -1s:
+1s:
* Matteo
* PengHui
* Dave
Thanks,
Matteo
--
Matteo Merli
On Thu, Jul 21, 2022 at 7:58 PM Dave Fisher wrote:
>
> Sorry I’m late to this discussion.
>
> I think that the motivation is correct. There is really quite a bit of
> activity around thi
Sorry I’m late to this discussion.
I think that the motivation is correct. There is really quite a bit of activity
around this issue. Let’s take extra efforts to engage extra time with commits
to confirm performance improvements.
Let’s particularly pay attention to threading.
+1
Regards,
Dave
+1
Penghui Li
On Jul 22, 2022, 02:37 +0800, Matteo Merli , wrote:
> ## Motivation
>
> The current implementation of the read cache in the Pulsar broker has largely
> remained unchanged for a long time, except for a few minor tweaks.
>
> While the implementation is stable and reasonably efficient f
## Motivation
The current implementation of the read cache in the Pulsar broker has largely
remained unchanged for a long time, except for a few minor tweaks.
While the implementation is stable and reasonably efficient for
typical workloads,
the overhead required for managing the cache evictions
