Tomasz Grabiec

On Tue, Nov 19, 2019 at 10:24 AM Shlomi Livne wrote: > @avikivity / @tgrabiec > - we have this patchset that reduces the > overall bootstrap time - it...

On Wed, Mar 18, 2020 at 2:07 AM Asias He wrote: > *@asias* commented on this pull request. > ------------------------------ > > In gms/gossiper.cc > : > > > -...

> > Our standard solution for this is to use cluster features. Not include newitems in the hash until the cluster is upgraded. It's more flexibly, itallows you to change...

> It seems to me that the partition index is not very useful when there is no clustering key. Maybe it should not be cached in this case. Why do...

to_data_query_result() is also called after reconciliation on coordinator side. The input frozen_mutation there may be the result of merging results from two replicas, and is not compacted. In this case,...

> In general I thought one of the guiding principles of the new tablet technique is that topology changes are trivial, and don't actually move tablets which are later moved...

> > The solution taken here is to not drain tablet replicas from replaced > > node during topology request but leave it to happen later after the > >...

On Tue, Feb 20, 2024 at 12:31 PM Gleb Natapov ***@***.***> wrote: > The solution taken here is to not drain tablet replicas from replaced > node during topology request...

On Tue, Feb 20, 2024 at 5:45 PM Gleb Natapov ***@***.***> wrote: > They care about replica pairing between base and view tables. Example: > > base: {n1, n2, n3}...

On Tue, Feb 20, 2024 at 6:19 PM Gleb Natapov ***@***.***> wrote: > That would make the get_replicas() API inconvenient. Instead of returning a > vector it would return vector,...