MTR fails and core dumps generated after cluster upgrade

1. install enterprise server 10.6.18-14 with Columnstore 23.10.2-1 on 3 nodes
2. start cluster
3. run `columnstore/setup` MTR test to full DB with data
4. upgrade server and Columnstore to 10.6.19-15 / 23.10.2-1 (no actual Columnstore upgrade)
("upgrade" process https://mariadb.com/docs/server/service-management/upgrades/enterprise-columnstore/multi-node/from-6-to-22-08/)
5. run other MTRs

result: almost all MTRs failed https://mdbe-ci-repo.mariadb.net/bb-logs/MariaDBEnterpriseProduction/10.6.19-15/alma_9_gcp/mtr-columnstore-multinode-reduced-test/2713/mtr.log , core dumps generated https://mdbe-ci-repo.mariadb.net/bb-logs/MariaDBEnterpriseProduction/10.6.19-15/alma_9_gcp/mtr-columnstore-multinode-reduced-test/2713/core_dumps.tar.gz

all logs:
https://mdbe-ci-repo.mariadb.net/bb-logs/MariaDBEnterpriseProduction/10.6.19-15/alma_9_gcp/mtr-columnstore-multinode-reduced-test/2713/