You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hugegraph.apache.org by GitBox <gi...@apache.org> on 2022/10/30 12:24:54 UTC
[GitHub] [incubator-hugegraph] gaeduf opened a new issue, #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
gaeduf opened a new issue, #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997
### Bug Type (问题类型)
gremlin (结果不合预期)
### Before submit
- [X] 我已经确认现有的 [Issues](https://github.com/hugegraph/hugegraph/issues) 与 [FAQ](https://hugegraph.github.io/hugegraph-doc/guides/faq.html) 中没有相同 / 重复问题
### Environment (环境信息)
- Server Version: v0.12.0
- Backend: RocksDB 单节点
- OS: 64 CPUs, 200 G RAM, Centos 7.
- Data Size: （顶点A） 2.6亿 vertices,（顶点A-B 关系） 1.9亿 edges （顶点B）0.33亿vertices
### Expected & Actual behavior (期望与实际表现)
如图所示：
这是顶点B的查询语句和结果 是正常的，
但是下图通过顶点A查询，却没有结果。
https://user-images.githubusercontent.com/113410236/198877308-c5490861-4660-4550-9fe8-f1b4062a3e2a.mp4
这是实操录像。
### Vertex/Edge example (问题点 / 边数据举例)
```javascript
这是查询语句，初步估计与数据无关
g.V('300007_9e47db0076d352672483d667751b88fa').bothE().otherV().path()
g.V('200209_7cee085c31b5b2fcc06427c5d3725ed6').bothE().otherV().path()
```
### Schema [VertexLabel, EdgeLabel, IndexLabel] (元数据结构)
```javascript
{
"vertices": [{
"label": "公司",
"input": {
"type": "file",
"path": "/sdb1/data/iflytek/hugegraph/hugegraph-loader-0.12.0/example/file/njaq/tb_mid_company_all_huge_day.txt",
"format": "TEXT",
"delimiter":"|++|",
"list_format":{
"start_symbol":"",
"elem_delimiter":"|",
"end_symbol":""
},
"header": ["id", "company", "day_id"],
"charset": "UTF-8",
"skipped_line": {
"regex": "(^#|^//).*"
}
},
"field_mapping":{
"id": "公司_id",
"company": "公司_公司",
"day_id": "公司_日期"
},
"id": "id",
"null_values": ["NULL", "null", ""]
}]
}
{
"vertices": [{
"label": "职业经历",
"input": {
"type": "file",
"path": "/sdb1/data/iflytek/hugegraph/hugegraph-loader-0.12.0/example/file/njaq/tb_mid_m_person_org_huge_day.txt",
"format": "TEXT",
"delimiter":"|++|",
"list_format":{
"start_symbol":"",
"elem_delimiter":"|",
"end_symbol":""
},
"header": ["entity_id","company","occupation_type","department","sub_department","title","begintime","endtime","name","dob","yob","sex","workphone","mobilephone","homephone","email","email_aft","address","username","password","graduation_school","abstract_flag","other","source"],
"charset": "UTF-8",
"skipped_line": {
"regex": "(^#|^//).*"
}
},
"field_mapping":{
"entity_id": "职业经历_id",
"company": "职业经历_公司",
"occupation_type": "职业经历_职业类型",
"department": "职业经历_部门",
"sub_department": "职业经历_子部门",
"title": "职业经历_职位",
"begintime": "职业经历_入职时间",
"endtime": "职业经历_离职时间",
"name": "职业经历_姓名",
"dob": "职业经历_生日",
"sex": "职业经历_性别",
"workphone": "职业经历_工作电话",
"mobilephone": "职业经历_移动电话",
"homephone": "职业经历_家庭电话",
"email": "职业经历_邮箱",
"email_aft": "职业经历_邮箱后缀",
"address": "职业经历_地址",
"username": "职业经历_用户名",
"password": "职业经历_明文密码",
"graduation_school": "职业经历_毕业学校",
"source": "职业经历_数据来源"
},
"id": "entity_id",
"ignored": ["yob","abstract_flag","other"]
}]
}
{
"edges": [ {
"label": "person_org_company_relation",
"source": ["entity_id"],
"target": ["company_id"],
"input": {
"type": "file",
"path": "/sdb1/data/iflytek/hugegraph/hugegraph-loader-0.12.0/example/file/njaq/tb_mid_rel_entity_company_huge_day.txt",
"format": "TEXT",
"delimiter":"|++|",
"list_format":{
"start_symbol":"",
"elem_delimiter":"|",
"end_symbol":""
},
"header": ["company_id","entity_id","company","occupation_type","department","sub_department","title","begintime","endtime","source"],
"charset": "UTF-8",
"skipped_line": {
"regex": "(^#|^//).*"
},
"date_format": "yyyyMMdd"
},
"field_mapping": {
"entity_id": "职业经历_id",
"company_id": "公司_id",
"company": "职业经历_公司",
"occupation_type": "职业经历_职业类型",
"department": "职业经历_部门",
"sub_department": "职业经历_子部门",
"title": "职业经历_职位",
"begintime": "职业经历_入职时间",
"endtime": "职业经历_离职时间",
"source": "relation_source"
}
}]
}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] javeme commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
javeme commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1302315601
请确保不存在悬空边（即可能不存在顶点A），可以执行`g.V(A_id)`进行验证。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1296248838
补充下：在我自己搭建的开发环境里，数据量很少时候不会出现这个问题。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1308067376
可以帮我远程看下吗？ @javeme 这个我想了三天三夜 都没想明白是怎么回事
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] javeme commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
javeme commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1313590833
导入时，需要确保一个文件中的每一行必须是同类数据，即数据的edgelable必须相同。
另外，`undefined`类型顶点表示不存在的顶点，有2种方式避免这种悬空边导致的问题：
1. 导入边时检查顶点必须存在，设置参数[`check_vertex=true`](https://hugegraph.apache.org/docs/clients/restful-api/edge/#222-%E5%88%9B%E5%BB%BA%E5%A4%9A%E6%9D%A1%E8%BE%B9)。
2. 或者后续导入缺少的顶点。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] javeme commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
javeme commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1312460679
> @javeme 对关于元数据使用语言的支持，请问是否对中文能完全支持，会不会出现不可预知的bug？
hugegraph完全支持中文字符的，实际上内部采用的是utf8编码，可以支持任意语言。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1312966500
> > @javeme 已经找到原因了，因为在边文件中，同一条边数据导入了三次，但是边名称不一样 这会导致上述那个问题。 另外： 我想问下，目前我们在定义边的时候 会指定 是A指向B 还是B指向A，在实践中，我发现 当边的struct.json文件 中 source 字段值填 taget的id值，target的字段 匹配的是source的id 时候，仍然可以正常连接（在useCustomizeStringId的时候），但是查询出来会出现 属性获取不到的情况。 为什么没有做校验呢？
>
> 没太明白“边名称不一样”的含义，能否给出具体示例schema与gremlin语句？
我描述的就是这种情况：
上图每两个顶点之间的关系都是三条边，其实只有一条边关系是正确的，之所以显示这种情况，是因为 这三条边关系都存放在一个文件里面去了，导致我们误认为 前面既然约束了 边的 起止顶点类型，不会出现这种 张冠李戴的情况，因而导入了三遍数据。但是事实上，边的起止顶点类型的定义没有任何约束作用（至少在使用自定义字段作为主键的情况下是这样的）。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1304973833
貌似不能用中文，已解决
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: [Question] gremlin 中一条边的两个出入点出发, 返回的路径不同
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1315265359
> 导入时，需要确保一个文件中的每一行必须是同类数据，即数据的edgelable必须相同。
>
> 另外，`undefined`类型顶点表示不存在的顶点，有2种方式避免这种悬空边导致的问题：
>
> 1. 导入边时检查顶点必须存在，设置参数[`check_vertex=true`](https://hugegraph.apache.org/docs/clients/restful-api/edge/#222-%E5%88%9B%E5%BB%BA%E5%A4%9A%E6%9D%A1%E8%BE%B9)。
> 2. 或者后续导入缺少的顶点。
对，其实 undefined 应该是查出来的这个点是不存在的。 实际情况是 如果用 g.V(id) 去查这个顶点 是存在的，只不过由于 前面边导入的有问题，因为在查出来的结果里面“显示是 undefined”。
目前来说，这个问题已经能弄清楚了。
感谢！
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1308056827
不是中文的问题，很奇怪，不知道为什么
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1308822656
> @gaeduf 可以尝试找出最小复现示例
能具体说说吗？
除此之外，我目前还有几个bug想反馈下。
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf closed issue #1997: [Question] gremlin 中一条边的两个出入点出发, 返回的路径不同
Posted by GitBox <gi...@apache.org>.
gaeduf closed issue #1997: [Question] gremlin 中一条边的两个出入点出发, 返回的路径不同
URL: https://github.com/apache/incubator-hugegraph/issues/1997
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1303632866
已验证 并存在
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1308067759
初步判断与数据量无关
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1312398586
@imbajin 可以关掉这个问题了，但是上面的疑问请解答下，感谢！
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] javeme commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
javeme commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1312460849
> @javeme 已经找到原因了，因为在边文件中，同一条边数据导入了三次，但是边名称不一样 这会导致上述那个问题。 另外： 我想问下，目前我们在定义边的时候 会指定 是A指向B 还是B指向A，在实践中，我发现 当边的struct.json文件 中 source 字段值填 taget的id值，target的字段 匹配的是source的id 时候，仍然可以正常连接（在useCustomizeStringId的时候），但是查询出来会出现 属性获取不到的情况。 为什么没有做校验呢？
没太明白“边名称不一样”的含义，能否给出具体示例schema与gremlin语句？
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1296244877
最近工期很紧，麻烦帮忙看看
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] javeme commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
javeme commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1308776091
@gaeduf 可以尝试找出最小复现示例
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1309992498
@javeme 已经找到原因了，因为在边文件中，同一条边数据导入了三次，但是边名称不一样 这会导致上述那个问题。
另外：
我想问下，目前我们在定义边的时候 会指定 是A指向B 还是B指向A，在实践中，我发现 当边的struct.json文件 中 source 字段值填 taget的id值，target的字段 匹配的是source的id 时候，仍然可以正常连接，但是查询出来会出现 属性获取不到的情况。
为什么没有做校验呢？
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-hugegraph] gaeduf commented on issue #1997: 在我的图数据库中：有一条数据关系：顶点 A --> 顶点B ，可以通过 g.V(B顶点id).bothE().otherV().path() 查询到，但是g.V(A顶点id).bothE().otherV().path()，查询不到。
Posted by GitBox <gi...@apache.org>.
gaeduf commented on issue #1997:
URL: https://github.com/apache/incubator-hugegraph/issues/1997#issuecomment-1309995424
@javeme 对关于元数据使用语言的支持，请问是否对中文能完全支持，会不会出现不可预知的bug？
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: dev-unsubscribe@hugegraph.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org