I am +1 on hardware accelerated compression support and I look forward to seeing the LZ4 and especially Zstd results. I do agree with Stefan that just like today where we fall back from native to pure java implementations when native code is not available we should maintain that clean automatic fallback when someone is running without access to QAT accelerators. I think that is addressed in the cep with "However, if the plugin version encounters an error, default compressor will handle the operation."
Thank you Shylaja for putting this CEP together! -Joey On Wed, Dec 17, 2025 at 1:55 AM Štefan Miklošovič <[email protected]> wrote: > +1 but I would like to see additional logic around what happens when > we fail to decompress with hardware. Either stop the node, or we > fallback to software de/compression, or we just log and update some > counter in the context of metrics etc. > > On Mon, Dec 15, 2025 at 9:55 PM Yifan Cai <[email protected]> wrote: > > > > +1 > > > > Thank you for putting all of this together. > > > > On Mon, Dec 15, 2025 at 11:43 AM Michael Shuler <[email protected]> > wrote: > >> > >> +1 > >> > >> On 12/15/25 11:01 AM, Kokoori, Shylaja wrote: > >> > Hello all, > >> > > >> > I am starting a new thread for the vote, based on feedback I received > on > >> > my earlier one. > >> > > >> > I would like to call a vote for CEP-49. > >> > > >> > Proposal: > >> > https://cwiki.apache.org/confluence/display/CASSANDRA/ > >> > CEP-49%3A+Hardware-accelerated+compression > >> > > >> > Discussion thread: > >> > https://lists.apache.org/thread/t766stj07rb9vgrg2cyyd5s7n3g2zdgt > >> > > >> > Voting guidelines > >> > https://www.apache.org/foundation/voting.html > >> > > >> > The vote will be open for 72 hours (longer if needed). > >> > Please, let me know if you need additional time to check the proposal > or > >> > have any questions. > >> > > >> > Thank you, > >> > Shylaja > >> > > >> >
