[Feature] refactor colocate

### Search before asking

- [x] I had searched in the [issues](https://github.com/apache/doris/issues?q=is%3Aissue) and found no similar issues.

### Description

当前colocate的限制：
1. 多租户部署时副本冗余：例如租户A需要JOIN租户B的表，租户C需要JOIN租户B的表，但是租户A不需要JOIN租户C的表，理论上我们只需要在租户A/C的机器上增加租户B的表的副本，租户A/C就可以在自己的机器上做ColocateJoin，但是当前的Colocate实现要求租户A/B/C的表必须在A/B/C三个租户的机器上都放置副本，导致副本冗余。
2. 分桶数限制：当前colocate的实现要求两个表的分桶数必须相同才能做ColocateJoin，当JOIN的表的数量级差别比较大时，难以设置统一的分桶数。

方案：
1. 支持租户粒度的colcoate，避免副本冗余问题。同时，单个租户的colocate失效不影响其他租户。

例如租户A/B/C上的3个表如下，t1与t2、t3与t2可以支持colocate join，没有副本冗余问题。
```
CREATE TABLE `t1` (
`id` bigint NOT NULL,
`m1` bigint NULL
) ENGINE=OLAP
duplicate KEY(`id`)
COMMENT 'OLAP'
DISTRIBUTED BY hash(id) BUCKETS 2
PROPERTIES (
"replication_allocation" = "tag.location.a: 1",
"colocate_group"="tag.location.a:coloc1"
);

CREATE TABLE `t2` (
`id` bigint NOT NULL,
`m2` bigint NULL
) ENGINE=OLAP
duplicate KEY(`id`)
COMMENT 'OLAP'
DISTRIBUTED BY hash(id) BUCKETS 2
PROPERTIES (
"replication_allocation" = "tag.location.a: 1,tag.location.b: 1,tag.location.c: 1",
"colocate_group"="tag.location.a:coloc1,tag.location.c:coloc3"
);

CREATE TABLE `t3` (
`id` bigint NOT NULL,
`m3` bigint NULL
) ENGINE=OLAP
duplicate KEY(`id`)
COMMENT 'OLAP'
DISTRIBUTED BY hash(id) BUCKETS 2
PROPERTIES (
"replication_allocation" = "tag.location.c: 1",
"colocate_group"="tag.location.c:coloc3"
);

```

2. 支持分桶数为倍数关系的colocate

例如下面的t4和t5的分桶数是t1的倍数，理论上t1/t4/t5也可以支持colocate join。(由于历史原因，我们内部的实现使用了两个表属性)

```
CREATE TABLE `t4` (
`id` bigint NOT NULL,
`m4` bigint NULL
) ENGINE=OLAP
duplicate KEY(`id`)
COMMENT 'OLAP'
DISTRIBUTED BY hash(id) BUCKETS 4
PROPERTIES (
"replication_allocation" = "tag.location.a: 1",
"colocate_slave"="tag.location.a:coloc1"
);

CREATE TABLE `t5` (
`id` bigint NOT NULL,
`m5` bigint NULL
) ENGINE=OLAP
duplicate KEY(`id`)
COMMENT 'OLAP'
DISTRIBUTED BY hash(id) BUCKETS 6
PROPERTIES (
"replication_allocation" = "tag.location.a: 1",
"colocate_slave"="tag.location.a:coloc1"
);

```

### Use case

_No response_

### Related issues

_No response_

### Are you willing to submit PR?

- [x] Yes I am willing to submit a PR!

### Code of Conduct

- [x] I agree to follow this project's [Code of Conduct](https://www.apache.org/foundation/policies/conduct)

No files or tests are named in the issue. Start by locating Doris's colocate-group handling and bucket-count validation, then identify the relevant tests. Done means tenant-scoped colocate joins avoid redundant replicas, tenant failures remain isolated, and compatible multiple bucket counts are supported.

Written by the indexing model from the issue text.