Re: Question on the migration tool

[Woonsan Ko <wo...@apache.org>]

Hi Phillip,

Disclaimer: I'm not an active user and haven't used the migration tool
yet. So, someone please chime in here with real expertise. (-;
Please see my comments below anyway.

On Wed, Feb 27, 2019 at 6:16 PM Suk Hyun Hwang <sx...@case.edu> wrote:
>
> Hello,
> Thank you so much for your reply.
>
> Is it really that easy to migrate an entire database over to S2Graph?
> The readme made it seem rather difficult to the untrained eye, which is why I sought out this godsend of a mail-list.

S2Graph is yet in "incubation" status. Volunteers here have been
trying to improve the documentation and others, but it doesn't seem
consistent enough.

> I have very little experience with databases, so I apologize in advance for the simple questions.

No worries. Whenever you have questions or find something to point out
or improve, feel free to let us know.

>
> I want to migrate an HBase database over to S2Graph. I'm guessing that the online migration tool does not apply to me here since RDBMS does not seem like a part of HBase.

I haven't used it for my life yet, but the documentation--from many
scattered pages--suggests MySQL is a prerequisite for migration:
[1] https://steamshon.gitbooks.io/s2graph-book/content/migration.html
     (The diagrams show RDBMS blocks.)
[2] https://steamshon.gitbooks.io/s2graph-book/content/build_loader.html
     ("dbUrl for s2graph core")
[3] https://libraries.io/github/SteamShon/s2graph
     ("...running following is required.... 2. mysqlsetup...)

I also noticed that some pages look outdated, but you seem to have
figured it out already (e.g, command line arguments).

>
> In that case, is all I have to do just use the loader.py file?
> Of course, I'll have to change all the args, which brings me to my next question.
>
> I'm guessing that I replace the first two args with my Hadoop/HBase configuration directory, use a simple name like "test" for the htable name, write down the address of the hadoop namenode.
>
> I'm not sure about the db_url, db_user, and db_password part.

See above.

> Will the migration from hbase to s2graph happen online as the script tries to log into my hbase? I thought I needed to reduce the HBase to a .csv file and give it as an input.

According to [1], you need "TSV file that follows the S2Graph bulk
load format" as input, and probably the loader tool will send those to
*both* RDBMS and S2Graph REST. That's my understanding from
architecture perspective. S2Graph experts, please correct me if I'm
wrong.

>
> Finally, if I make the auto_create_edge field false, how do I create the edges myself?

I vaguely assume you have to create those through S2Graph REST APIs...

Anyway, hope someone shed a light here. ;-)

Woonsan

>
> Thanks you so much for your help, and I apologize again for my embarrassing lack of knowledge.
>
> Sincerely,
> Phillip
>
> On Wed, Feb 27, 2019 at 4:21 AM Hwansung Yu <ra...@apache.org> wrote:
>>
>> Hi, Phillip Hwang
>>
>> Welcome to S2Graph.
>> S2Graph is an HBase based graph database.
>> (Of course, other storage is also supported.)
>> In other words, if you use S2graph to store data, you can also graph traversal on HBase.
>> To migrate, you must make the existing data HFile and store it in HBase.
>> This can be done using the code below.
>> https://github.com/apache/incubator-s2graph/tree/master/loader
>>
>> To give a little more explanation,
>> The original file you want to migrate exists in the form of CSV, which you can migrate via loader.
>>
>> You can find more information in the following documents:
>> https://steamshon.gitbooks.io/s2graph-book/content/migration.html
>>
>> Thanks for your interest.
>>
>> Regards.
>>
>> 2019년 2월 26일 (화) 오전 6:29, Suk Hyun Hwang <sx...@case.edu>님이 작성:
>>>
>>> Hello,
>>> I am a current Masters student at Case Western Reserve University trying to graduate. My PI wants me to implement S2Graph into my research lab after looking at all the great things S2Graph has to offer. He is hinting at the extremely enticing fruit of graduation and being set free!!!
>>>
>>> I'm having trouble figuring out how to use the migration tool on the readme. It seems to be a very simple process to migrate data over from HBase to S2Graph using a built-in loader, but I'm not sure how to proceed. If anybody could help me out with this, I'd be extremely grateful.
>>> Thank you.
>>> Sincerely,
>>> Phillip Hwang
>>>
>>> 안녕하세요,
>>> 현재 미국 Case Western Reserve University 에서 석사과정 유학중인 랩노예 황석현이라고 합니다. 2017년 여름 판교 SK플래닛 캠퍼스 ML Lab 에서 인턴을 했었는데요, 11번가랑 신세계랑의 합체 루머, 11번가 분사 등등 혼란스러운 틈을 타서 다음카카오쪽으로 친한 매니저님들이 많이 넘어가셨습니다. 다음카카오에 대한 좋은 이야기 많이 듣고 있었던 참에 S2Graph 가 이쪽거라는것을 듣고 참 다행이라는 생각 많이 했습니다!
>>>
>>> 담당 교수님께서 우리 랩 디비를 Graph Database 로 바꾸라고 하셨는데, 졸업시켜주시겠다는 눈치를 아주 사알짝 주셔서 조심스럽게 자유와 졸업, 그리고 군대...(ㅜㅜ) 에 대한 작은 희망을 품고 있습니다. 제 해방을 도와주시는 분들 모두 이번 5월달, 한국 도착해서 치맥 시원하게 한턱 쏘겠습니다!
>>>
>>> README를 보면 간단하게 HBase 에서 S2Graph 로 sbt loader 를 이용해서 옮길 수 있다고 하는데요, 이 부분이 잘 이해가 안됩니다. 어떻게 옮기면 되는거죠?
>>>
>>> 감사합니다.
>>> 석현올림

Re: Question on the migration tool

[Suk Hyun Hwang <sx...@case.edu>]

Oh my god, you have no idea how helpful this is.
Thank you so much! I have been trying to process all this information over
the past few days and I think I got the gist of a lot of things. I would
have never been able to figure it out without your help. I'm so impressed
to see such a healthy and supportive s2graph user community.

Yes, I will definitely send you some sample data soon after discussing this
with my professor.

By the way, if I were to want to keep HBase on the side, how would I go
about automatically updating S2Graph when data in HBase changes, added, or
deleted?

Thank you again, and I hope you all have a wonderful week.
Sincerely,
Phillip

On Tue, Mar 5, 2019 at 1:35 AM DO YUNG YOON <sh...@gmail.com> wrote:

> Hi Phillip Hwang.
> Thanks for asking the questions, and sorry for the late reply.
>
> Like you and Woonsan pointed out, the documentation links above are
> outdated and not correct.
>
> Please take a look at
> https://github.com/apache/incubator-s2graph/tree/master/s2jobs to see how
> to migrate in general(loader project has been deprecated).
>
> Also, here is the simple example, which download the movielens dataset
> then, upload into local s2graph.
> https://github.com/apache/incubator-s2graph/tree/master/example/movielens#running-this-example
>
> I see the process is not documented so please ask any questions if you are
> having trouble.
>
> It would be also good to explain how the data is stored in your existing
> HBase.
> If you can share the sample data, then I can go through the bulkloader
> with sample data then see if it works.
>
> So in general, migration would be like this.
>
> 1. dump data which already stored in the existing system
> - Based on your comment, I am guessing your data reside on HBase already.
> then It would be necessary to dump data into Spark DataFrame(or any file
> format which can be translated into Edge/Vertex)
> -
> https://github.com/apache/incubator-s2graph/blob/master/s2jobs/src/main/scala/org/apache/s2graph/s2jobs/serde/reader/S2GraphCellReader.scala
> shows how to implement Reader which translate your old data to GraphElement.
> - I think the way we should go is asking users who want to migrate to
> provide the implementation of GraphElementReadable, but I don't think we
> are not yet there. I am more than happy to help out and clean up the
> current mess.
>
> 2. create the schema.
> - S2Graph require the user to create Service/ServiceColumn/Label schema
> which define metadata for Vertex/Edge.
> - how to create the schema in the system, here is a very brief guide
> https://github.com/apache/incubator-s2graph#your-first-graph.
>
> 3. build HFile using spark.
> - once finishing with step 1, s2graph expect the vertex/edge data should
> exist on HDFS(or Spark DataFrame).
> -
> https://github.com/apache/incubator-s2graph/blob/master/s2jobs/src/main/scala/org/apache/s2graph/s2jobs/serde/writer/RowDataFrameWriter.scala
> shows how to implement Writer which translate your dumped data to
> GraphElement.
> - same as 1.
>
> 4. distcp HFile to a production cluster.
> - use hadoop distcp command to copy HFile from 3 to HBase cluster.
> - then issue LoadIncrementalHFiles to HBase cluster.
>
> If your existing data is relatively small(say 1~10 million vertex/edges)
> then it would be easier if you just set up s2http server then send HTTP
> POST request(I think it would only take few minutes to process ~10 million
> data if the setting is correct).
>
> Finally, answers in line.
>
> > I want to migrate an HBase database over to S2Graph. I'm guessing that
> the online migration tool does not apply to me here since RDBMS does not
> seem like a part of HBase.
>
> users need to setup s2graph with running RDBMS to store only schema, not
> the actual vertex/edge data.
> the default meta storage use h2 with local file system. if you run the
> movielens example mentioned above, it setup local s2http server with h2.
>
> db_url = any valid jdbc connection info to connect to RDBMS for meta
> storage. ex) "jdbc://postgresql://host:port/database"
> db_password = password for database
>
> Let me know if you need more explanation.
>
> > Finally, if I make the auto_create_edge field false, how do I create the
> edges myself?
>
> This options is for the feature that create inverse edge automatically(not
> the edge itself, the name sounds confusing).
>
> So lets say user insert one edge as follow.
>
> from -> direction = out -> to
> user_1 -> liked -> movie_1
>
> then if auto_create_edge is set, then loader automatically create
> following edge.
>
> from <- direction = in <- to
> movie_1 <- liked <- user_1
>
> if you need to start traverse from movie_1, then you should insert two
> edges manually, or set auto_create_edge.
>
> Hope this is helpful.
>
>
>
>
>
> On Tue, Mar 5, 2019 at 1:01 AM Woonsan Ko <wo...@apache.org> wrote:
>
>> Hi Phillip,
>>
>> Disclaimer: I'm not an active user and haven't used the migration tool
>> yet. So, someone please chime in here with real expertise. (-;
>> Please see my comments below anyway.
>>
>> On Wed, Feb 27, 2019 at 6:16 PM Suk Hyun Hwang <sx...@case.edu> wrote:
>> >
>> > Hello,
>> > Thank you so much for your reply.
>> >
>> > Is it really that easy to migrate an entire database over to S2Graph?
>> > The readme made it seem rather difficult to the untrained eye, which is
>> why I sought out this godsend of a mail-list.
>>
>> S2Graph is yet in "incubation" status. Volunteers here have been
>> trying to improve the documentation and others, but it doesn't seem
>> consistent enough.
>>
>> > I have very little experience with databases, so I apologize in advance
>> for the simple questions.
>>
>> No worries. Whenever you have questions or find something to point out
>> or improve, feel free to let us know.
>>
>> >
>> > I want to migrate an HBase database over to S2Graph. I'm guessing that
>> the online migration tool does not apply to me here since RDBMS does not
>> seem like a part of HBase.
>>
>> I haven't used it for my life yet, but the documentation--from many
>> scattered pages--suggests MySQL is a prerequisite for migration:
>> [1] https://steamshon.gitbooks.io/s2graph-book/content/migration.html
>>      (The diagrams show RDBMS blocks.)
>> [2] https://steamshon.gitbooks.io/s2graph-book/content/build_loader.html
>>      ("dbUrl for s2graph core")
>> [3] https://libraries.io/github/SteamShon/s2graph
>>      ("...running following is required.... 2. mysqlsetup...)
>>
>> I also noticed that some pages look outdated, but you seem to have
>> figured it out already (e.g, command line arguments).
>>
>> >
>> > In that case, is all I have to do just use the loader.py file?
>> > Of course, I'll have to change all the args, which brings me to my next
>> question.
>> >
>> > I'm guessing that I replace the first two args with my Hadoop/HBase
>> configuration directory, use a simple name like "test" for the htable name,
>> write down the address of the hadoop namenode.
>> >
>> > I'm not sure about the db_url, db_user, and db_password part.
>>
>> See above.
>>
>> > Will the migration from hbase to s2graph happen online as the script
>> tries to log into my hbase? I thought I needed to reduce the HBase to a
>> .csv file and give it as an input.
>>
>> According to [1], you need "TSV file that follows the S2Graph bulk
>> load format" as input, and probably the loader tool will send those to
>> *both* RDBMS and S2Graph REST. That's my understanding from
>> architecture perspective. S2Graph experts, please correct me if I'm
>> wrong.
>>
>> >
>> > Finally, if I make the auto_create_edge field false, how do I create
>> the edges myself?
>>
>> I vaguely assume you have to create those through S2Graph REST APIs...
>>
>> Anyway, hope someone shed a light here. ;-)
>>
>> Woonsan
>>
>> >
>> > Thanks you so much for your help, and I apologize again for my
>> embarrassing lack of knowledge.
>> >
>> > Sincerely,
>> > Phillip
>> >
>> > On Wed, Feb 27, 2019 at 4:21 AM Hwansung Yu <ra...@apache.org> wrote:
>> >>
>> >> Hi, Phillip Hwang
>> >>
>> >> Welcome to S2Graph.
>> >> S2Graph is an HBase based graph database.
>> >> (Of course, other storage is also supported.)
>> >> In other words, if you use S2graph to store data, you can also graph
>> traversal on HBase.
>> >> To migrate, you must make the existing data HFile and store it in
>> HBase.
>> >> This can be done using the code below.
>> >> https://github.com/apache/incubator-s2graph/tree/master/loader
>> >>
>> >> To give a little more explanation,
>> >> The original file you want to migrate exists in the form of CSV, which
>> you can migrate via loader.
>> >>
>> >> You can find more information in the following documents:
>> >> https://steamshon.gitbooks.io/s2graph-book/content/migration.html
>> >>
>> >> Thanks for your interest.
>> >>
>> >> Regards.
>> >>
>> >> 2019년 2월 26일 (화) 오전 6:29, Suk Hyun Hwang <sx...@case.edu>님이 작성:
>> >>>
>> >>> Hello,
>> >>> I am a current Masters student at Case Western Reserve University
>> trying to graduate. My PI wants me to implement S2Graph into my research
>> lab after looking at all the great things S2Graph has to offer. He is
>> hinting at the extremely enticing fruit of graduation and being set free!!!
>> >>>
>> >>> I'm having trouble figuring out how to use the migration tool on the
>> readme. It seems to be a very simple process to migrate data over from
>> HBase to S2Graph using a built-in loader, but I'm not sure how to proceed.
>> If anybody could help me out with this, I'd be extremely grateful.
>> >>> Thank you.
>> >>> Sincerely,
>> >>> Phillip Hwang
>> >>>
>> >>> 안녕하세요,
>> >>> 현재 미국 Case Western Reserve University 에서 석사과정 유학중인 랩노예 황석현이라고 합니다.
>> 2017년 여름 판교 SK플래닛 캠퍼스 ML Lab 에서 인턴을 했었는데요, 11번가랑 신세계랑의 합체 루머, 11번가 분사 등등
>> 혼란스러운 틈을 타서 다음카카오쪽으로 친한 매니저님들이 많이 넘어가셨습니다. 다음카카오에 대한 좋은 이야기 많이 듣고 있었던 참에
>> S2Graph 가 이쪽거라는것을 듣고 참 다행이라는 생각 많이 했습니다!
>> >>>
>> >>> 담당 교수님께서 우리 랩 디비를 Graph Database 로 바꾸라고 하셨는데, 졸업시켜주시겠다는 눈치를 아주 사알짝
>> 주셔서 조심스럽게 자유와 졸업, 그리고 군대...(ㅜㅜ) 에 대한 작은 희망을 품고 있습니다. 제 해방을 도와주시는 분들 모두 이번
>> 5월달, 한국 도착해서 치맥 시원하게 한턱 쏘겠습니다!
>> >>>
>> >>> README를 보면 간단하게 HBase 에서 S2Graph 로 sbt loader 를 이용해서 옮길 수 있다고 하는데요, 이
>> 부분이 잘 이해가 안됩니다. 어떻게 옮기면 되는거죠?
>> >>>
>> >>> 감사합니다.
>> >>> 석현올림
>>
>

Re: Question on the migration tool

[DO YUNG YOON <sh...@gmail.com>]

Hi Phillip Hwang.
Thanks for asking the questions, and sorry for the late reply.

Like you and Woonsan pointed out, the documentation links above are
outdated and not correct.

Please take a look at
https://github.com/apache/incubator-s2graph/tree/master/s2jobs to see how
to migrate in general(loader project has been deprecated).

Also, here is the simple example, which download the movielens dataset
then, upload into local s2graph.
https://github.com/apache/incubator-s2graph/tree/master/example/movielens#running-this-example

I see the process is not documented so please ask any questions if you are
having trouble.

It would be also good to explain how the data is stored in your existing
HBase.
If you can share the sample data, then I can go through the bulkloader with
sample data then see if it works.

So in general, migration would be like this.

1. dump data which already stored in the existing system
- Based on your comment, I am guessing your data reside on HBase already.
then It would be necessary to dump data into Spark DataFrame(or any file
format which can be translated into Edge/Vertex)
-
https://github.com/apache/incubator-s2graph/blob/master/s2jobs/src/main/scala/org/apache/s2graph/s2jobs/serde/reader/S2GraphCellReader.scala
shows how to implement Reader which translate your old data to GraphElement.
- I think the way we should go is asking users who want to migrate to
provide the implementation of GraphElementReadable, but I don't think we
are not yet there. I am more than happy to help out and clean up the
current mess.

2. create the schema.
- S2Graph require the user to create Service/ServiceColumn/Label schema
which define metadata for Vertex/Edge.
- how to create the schema in the system, here is a very brief guide
https://github.com/apache/incubator-s2graph#your-first-graph.

3. build HFile using spark.
- once finishing with step 1, s2graph expect the vertex/edge data should
exist on HDFS(or Spark DataFrame).
-
https://github.com/apache/incubator-s2graph/blob/master/s2jobs/src/main/scala/org/apache/s2graph/s2jobs/serde/writer/RowDataFrameWriter.scala
shows how to implement Writer which translate your dumped data to
GraphElement.
- same as 1.

4. distcp HFile to a production cluster.
- use hadoop distcp command to copy HFile from 3 to HBase cluster.
- then issue LoadIncrementalHFiles to HBase cluster.

If your existing data is relatively small(say 1~10 million vertex/edges)
then it would be easier if you just set up s2http server then send HTTP
POST request(I think it would only take few minutes to process ~10 million
data if the setting is correct).

Finally, answers in line.

> I want to migrate an HBase database over to S2Graph. I'm guessing that
the online migration tool does not apply to me here since RDBMS does not
seem like a part of HBase.

users need to setup s2graph with running RDBMS to store only schema, not
the actual vertex/edge data.
the default meta storage use h2 with local file system. if you run the
movielens example mentioned above, it setup local s2http server with h2.

db_url = any valid jdbc connection info to connect to RDBMS for meta
storage. ex) "jdbc://postgresql://host:port/database"
db_password = password for database

Let me know if you need more explanation.

> Finally, if I make the auto_create_edge field false, how do I create the
edges myself?

This options is for the feature that create inverse edge automatically(not
the edge itself, the name sounds confusing).

So lets say user insert one edge as follow.

from -> direction = out -> to
user_1 -> liked -> movie_1

then if auto_create_edge is set, then loader automatically create following
edge.

from <- direction = in <- to
movie_1 <- liked <- user_1

if you need to start traverse from movie_1, then you should insert two
edges manually, or set auto_create_edge.

Hope this is helpful.





On Tue, Mar 5, 2019 at 1:01 AM Woonsan Ko <wo...@apache.org> wrote:

> Hi Phillip,
>
> Disclaimer: I'm not an active user and haven't used the migration tool
> yet. So, someone please chime in here with real expertise. (-;
> Please see my comments below anyway.
>
> On Wed, Feb 27, 2019 at 6:16 PM Suk Hyun Hwang <sx...@case.edu> wrote:
> >
> > Hello,
> > Thank you so much for your reply.
> >
> > Is it really that easy to migrate an entire database over to S2Graph?
> > The readme made it seem rather difficult to the untrained eye, which is
> why I sought out this godsend of a mail-list.
>
> S2Graph is yet in "incubation" status. Volunteers here have been
> trying to improve the documentation and others, but it doesn't seem
> consistent enough.
>
> > I have very little experience with databases, so I apologize in advance
> for the simple questions.
>
> No worries. Whenever you have questions or find something to point out
> or improve, feel free to let us know.
>
> >
> > I want to migrate an HBase database over to S2Graph. I'm guessing that
> the online migration tool does not apply to me here since RDBMS does not
> seem like a part of HBase.
>
> I haven't used it for my life yet, but the documentation--from many
> scattered pages--suggests MySQL is a prerequisite for migration:
> [1] https://steamshon.gitbooks.io/s2graph-book/content/migration.html
>      (The diagrams show RDBMS blocks.)
> [2] https://steamshon.gitbooks.io/s2graph-book/content/build_loader.html
>      ("dbUrl for s2graph core")
> [3] https://libraries.io/github/SteamShon/s2graph
>      ("...running following is required.... 2. mysqlsetup...)
>
> I also noticed that some pages look outdated, but you seem to have
> figured it out already (e.g, command line arguments).
>
> >
> > In that case, is all I have to do just use the loader.py file?
> > Of course, I'll have to change all the args, which brings me to my next
> question.
> >
> > I'm guessing that I replace the first two args with my Hadoop/HBase
> configuration directory, use a simple name like "test" for the htable name,
> write down the address of the hadoop namenode.
> >
> > I'm not sure about the db_url, db_user, and db_password part.
>
> See above.
>
> > Will the migration from hbase to s2graph happen online as the script
> tries to log into my hbase? I thought I needed to reduce the HBase to a
> .csv file and give it as an input.
>
> According to [1], you need "TSV file that follows the S2Graph bulk
> load format" as input, and probably the loader tool will send those to
> *both* RDBMS and S2Graph REST. That's my understanding from
> architecture perspective. S2Graph experts, please correct me if I'm
> wrong.
>
> >
> > Finally, if I make the auto_create_edge field false, how do I create the
> edges myself?
>
> I vaguely assume you have to create those through S2Graph REST APIs...
>
> Anyway, hope someone shed a light here. ;-)
>
> Woonsan
>
> >
> > Thanks you so much for your help, and I apologize again for my
> embarrassing lack of knowledge.
> >
> > Sincerely,
> > Phillip
> >
> > On Wed, Feb 27, 2019 at 4:21 AM Hwansung Yu <ra...@apache.org> wrote:
> >>
> >> Hi, Phillip Hwang
> >>
> >> Welcome to S2Graph.
> >> S2Graph is an HBase based graph database.
> >> (Of course, other storage is also supported.)
> >> In other words, if you use S2graph to store data, you can also graph
> traversal on HBase.
> >> To migrate, you must make the existing data HFile and store it in HBase.
> >> This can be done using the code below.
> >> https://github.com/apache/incubator-s2graph/tree/master/loader
> >>
> >> To give a little more explanation,
> >> The original file you want to migrate exists in the form of CSV, which
> you can migrate via loader.
> >>
> >> You can find more information in the following documents:
> >> https://steamshon.gitbooks.io/s2graph-book/content/migration.html
> >>
> >> Thanks for your interest.
> >>
> >> Regards.
> >>
> >> 2019년 2월 26일 (화) 오전 6:29, Suk Hyun Hwang <sx...@case.edu>님이 작성:
> >>>
> >>> Hello,
> >>> I am a current Masters student at Case Western Reserve University
> trying to graduate. My PI wants me to implement S2Graph into my research
> lab after looking at all the great things S2Graph has to offer. He is
> hinting at the extremely enticing fruit of graduation and being set free!!!
> >>>
> >>> I'm having trouble figuring out how to use the migration tool on the
> readme. It seems to be a very simple process to migrate data over from
> HBase to S2Graph using a built-in loader, but I'm not sure how to proceed.
> If anybody could help me out with this, I'd be extremely grateful.
> >>> Thank you.
> >>> Sincerely,
> >>> Phillip Hwang
> >>>
> >>> 안녕하세요,
> >>> 현재 미국 Case Western Reserve University 에서 석사과정 유학중인 랩노예 황석현이라고 합니다.
> 2017년 여름 판교 SK플래닛 캠퍼스 ML Lab 에서 인턴을 했었는데요, 11번가랑 신세계랑의 합체 루머, 11번가 분사 등등
> 혼란스러운 틈을 타서 다음카카오쪽으로 친한 매니저님들이 많이 넘어가셨습니다. 다음카카오에 대한 좋은 이야기 많이 듣고 있었던 참에
> S2Graph 가 이쪽거라는것을 듣고 참 다행이라는 생각 많이 했습니다!
> >>>
> >>> 담당 교수님께서 우리 랩 디비를 Graph Database 로 바꾸라고 하셨는데, 졸업시켜주시겠다는 눈치를 아주 사알짝 주셔서
> 조심스럽게 자유와 졸업, 그리고 군대...(ㅜㅜ) 에 대한 작은 희망을 품고 있습니다. 제 해방을 도와주시는 분들 모두 이번 5월달,
> 한국 도착해서 치맥 시원하게 한턱 쏘겠습니다!
> >>>
> >>> README를 보면 간단하게 HBase 에서 S2Graph 로 sbt loader 를 이용해서 옮길 수 있다고 하는데요, 이
> 부분이 잘 이해가 안됩니다. 어떻게 옮기면 되는거죠?
> >>>
> >>> 감사합니다.
> >>> 석현올림
>