- In sql tests (test_limit) we encountered a bug,
when two different queries had the same hash
for executable subtree. We marked teset as xfail.
- The bug was caused by wrong calculation of hash
in sbroad. This was solved.
- Update sbroad submodule to commit with fix
- Remove xfail mark from test

See https://git.picodata.io/picodata/picodata/sbroad/-/issues/837

Needed because $CI_COMMIT_BRANCH is not available for tags

Note that now we don't have any (is this right?) tests that use pico.cas
which should be considering we plan to get rid of it in the near future.

Also we don't use `password` for cas operations, because the stored
procedure can only be called by the superuser and only the access
control checks rely on the `as_user` parameter.

See doc comments for RouteBuilder::register, but TL;DR it's never safe
to wrap a yielding function in a FnMut.
The user should instead do the regular internal mutability thing.

Up to this patch `instance_uuid` and `replicaset_uuid` were generated
using name-based UUID v3. So all instances named `i1` in all clusters
had the same UUID.

With this patch UUIDs become random v4. Particular UUID values are
generated in two locations:

1. `start_boot` (src/lib.rs)
2. `handle_join_request_and_wait` -> `build_instance` (src/rpc/join.rs)

Function `Instance::new()` is removed in favor of explicit struct
construction.

The main purpose is to reduce the usage of `Instance::new()` function.
It's going to be removed soon. No behavior changes are introduced yet.

This reverts commit 2924f112.

Kurdakov Alexander authored 6 months ago and Yaroslav Dynnikov committed 5 months ago

Do not ban admin console via unix socket
after several failed attempts via `picodata connect`.

godzie44 authored 6 months ago and Georgy Moshkin committed 5 months ago

Closes #916

godzie44 authored 6 months ago and Georgy Moshkin committed 5 months ago

Closes #664