[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15325867#comment-15325867 ] DOYUNG YOON commented on S2GRAPH-70: [~jongwook] PR works like charm. I just merged your PR into master. Thanks for the great contribution. > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > Fix For: 0.1.0 > > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15325866#comment-15325866 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user asfgit closed the pull request at: https://github.com/apache/incubator-s2graph/pull/57 > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > Fix For: 0.1.0 > > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15324516#comment-15324516 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user daewon commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 LGTM +1 > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > Fix For: 0.1.0 > > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15321673#comment-15321673 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user SteamShon commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 Never knew [FOSS exception](https://www.mysql.com/about/legal/licensing/foss-exception/) before. Anyway I think PR is great and I am going to merge this tonight. It would be better not to worry about license here and we can create separate issue dedicate on license compatibility checks. > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320669#comment-15320669 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user jongwook commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 We can also consider [MariaDB Connector/J](https://downloads.mariadb.org/connector-java/) which is LGPL licensed. > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320654#comment-15320654 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user jongwook commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 Oracle offers [FOSS exception](https://www.mysql.com/about/legal/licensing/foss-exception/) for mysql-connector's GPL license, where Apache Software License is one of them, so we might be safe after all. Should we get some legal advices before proceeding? > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320638#comment-15320638 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user jongwook commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 If that's the cause, I guess we should exclude mysql-connector-java and document it that way .. > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320632#comment-15320632 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user SteamShon commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 I think mysql connector has GPL lisence so was wondering if we should remove mysql connector this time and ask user to download it. I think there are many other project follow this way because of license issue > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[
https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320599#comment-15320599
]
ASF GitHub Bot commented on S2GRAPH-70:
---
Github user jongwook commented on a diff in the pull request:
https://github.com/apache/incubator-s2graph/pull/57#discussion_r66261184
--- Diff: s2core/build.sbt ---
@@ -26,15 +26,18 @@ libraryDependencies ++= Seq(
"com.typesafe" % "config" % "1.2.1",
"com.typesafe.play" %% "play-json" % Common.playVersion,
"com.typesafe.akka" %% "akka-actor" % "2.3.4",
- "org.apache.hbase" % "hbase-client" % Common.hbaseVersion excludeAll
ExclusionRule(organization = "org.slf4j"),
- "org.apache.hbase" % "hbase-common" % Common.hbaseVersion excludeAll
ExclusionRule(organization = "org.slf4j"),
- "org.apache.hbase" % "hbase-server" % Common.hbaseVersion
excludeAll(ExclusionRule(organization = "org.slf4j"),
ExclusionRule(organization = "com.google.protobuf")),
- "org.apache.hadoop" % "hadoop-common" % Common.hadoopVersion excludeAll
ExclusionRule(organization = "org.slf4j"),
+ "com.google.guava" % "guava" % "12.0.1" force(), // use this old version
of guava to avoid incompatibility
+ "org.apache.hbase" % "hbase-client" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.hbase" % "hbase-common" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.hbase" % "hbase-server" % Common.hbaseVersion
exclude("org.slf4j", "*") exclude("com.google.protobuf", "*"),
+ "org.apache.hbase" % "hbase-hadoop-compat" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.hbase" % "hbase-hadoop2-compat" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.kafka" % "kafka-clients" % "0.8.2.0" exclude("org.slf4j",
"*") exclude("com.sun.jdmk", "*") exclude("com.sun.jmx", "*")
exclude("javax.jms", "*"),
"commons-pool" % "commons-pool" % "1.6",
"org.scalatest" %% "scalatest" % "2.2.4" % "test",
"org.scalikejdbc" %% "scalikejdbc" % "2.1.+",
"mysql" % "mysql-connector-java" % "5.1.28",
--- End diff --
Well, technically mysql-connector-java.jar Is required only if the user
needs to use mysql as metastore, but I guess it's easier to just include
mysql-connector in the UX perspective because they can just change JDBC url in
application.conf without downloading jar from Oracle or Maven central manually.
> Automate the process of building a distribution package
> ---
>
> Key: S2GRAPH-70
> URL: https://issues.apache.org/jira/browse/S2GRAPH-70
> Project: S2Graph
> Issue Type: Improvement
>Reporter: Jong Wook Kim
>
> Since the Vagrant image (which is the only documented way to trying out
> S2Graph) is outdated, and also as a prerequisite of making a release, it is
> desirable to be able to readily make a distribution package. The goal is to
> be able for an end-user to download & untar the package, and run a shell
> script to start off s2graph and all its dependencies.
> AFAIK S2graph requires mysql for storing the metadata and HBase for the
> actual data. Both could be contained in the package if:
> - we run HBase server that is in the hbase-server jar, running without HDFS,
> saving data in the local filesystem instead, and
> - instead of the full MySQL server, use an embedded database like derby or h2
> to store the metadata.
> It would also require some technical consideration about how exactly we
> should layout the files and what toolchain we should use to build the
> package, so I think adding child issues for each subtask would be appropriate.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320584#comment-15320584 ] ASF GitHub Bot commented on S2GRAPH-70: --- Github user SteamShon commented on the issue: https://github.com/apache/incubator-s2graph/pull/57 looks good to me! +1 > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[
https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15320581#comment-15320581
]
ASF GitHub Bot commented on S2GRAPH-70:
---
Github user SteamShon commented on a diff in the pull request:
https://github.com/apache/incubator-s2graph/pull/57#discussion_r66258989
--- Diff: s2core/build.sbt ---
@@ -26,15 +26,18 @@ libraryDependencies ++= Seq(
"com.typesafe" % "config" % "1.2.1",
"com.typesafe.play" %% "play-json" % Common.playVersion,
"com.typesafe.akka" %% "akka-actor" % "2.3.4",
- "org.apache.hbase" % "hbase-client" % Common.hbaseVersion excludeAll
ExclusionRule(organization = "org.slf4j"),
- "org.apache.hbase" % "hbase-common" % Common.hbaseVersion excludeAll
ExclusionRule(organization = "org.slf4j"),
- "org.apache.hbase" % "hbase-server" % Common.hbaseVersion
excludeAll(ExclusionRule(organization = "org.slf4j"),
ExclusionRule(organization = "com.google.protobuf")),
- "org.apache.hadoop" % "hadoop-common" % Common.hadoopVersion excludeAll
ExclusionRule(organization = "org.slf4j"),
+ "com.google.guava" % "guava" % "12.0.1" force(), // use this old version
of guava to avoid incompatibility
+ "org.apache.hbase" % "hbase-client" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.hbase" % "hbase-common" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.hbase" % "hbase-server" % Common.hbaseVersion
exclude("org.slf4j", "*") exclude("com.google.protobuf", "*"),
+ "org.apache.hbase" % "hbase-hadoop-compat" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.hbase" % "hbase-hadoop2-compat" % Common.hbaseVersion
exclude("org.slf4j", "*"),
+ "org.apache.kafka" % "kafka-clients" % "0.8.2.0" exclude("org.slf4j",
"*") exclude("com.sun.jdmk", "*") exclude("com.sun.jmx", "*")
exclude("javax.jms", "*"),
"commons-pool" % "commons-pool" % "1.6",
"org.scalatest" %% "scalatest" % "2.2.4" % "test",
"org.scalikejdbc" %% "scalikejdbc" % "2.1.+",
"mysql" % "mysql-connector-java" % "5.1.28",
--- End diff --
Just wondering, do we need both of mysql-connector-java and h2?
> Automate the process of building a distribution package
> ---
>
> Key: S2GRAPH-70
> URL: https://issues.apache.org/jira/browse/S2GRAPH-70
> Project: S2Graph
> Issue Type: Improvement
>Reporter: Jong Wook Kim
>
> Since the Vagrant image (which is the only documented way to trying out
> S2Graph) is outdated, and also as a prerequisite of making a release, it is
> desirable to be able to readily make a distribution package. The goal is to
> be able for an end-user to download & untar the package, and run a shell
> script to start off s2graph and all its dependencies.
> AFAIK S2graph requires mysql for storing the metadata and HBase for the
> actual data. Both could be contained in the package if:
> - we run HBase server that is in the hbase-server jar, running without HDFS,
> saving data in the local filesystem instead, and
> - instead of the full MySQL server, use an embedded database like derby or h2
> to store the metadata.
> It would also require some technical consideration about how exactly we
> should layout the files and what toolchain we should use to build the
> package, so I think adding child issues for each subtask would be appropriate.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (S2GRAPH-70) Automate the process of building a distribution package
[ https://issues.apache.org/jira/browse/S2GRAPH-70?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15317444#comment-15317444 ] ASF GitHub Bot commented on S2GRAPH-70: --- GitHub user jongwook opened a pull request: https://github.com/apache/incubator-s2graph/pull/57 [S2GRAPH-70] Automate the process of building a distribution package This pull request resolves [S2GRAPH-70](https://issues.apache.org/jira/browse/S2GRAPH-70). It augments the SBT command `activator package` to create a package for distribution at the directory `target/deploy/`. Within the package, the command `bin/start-s2graph.sh` will launch the S2Graph server along with a standalone HBase server as the data storage and H2 as the metastore. For details, refer to the subtasks [S2GRAPH-73](https://issues.apache.org/jira/browse/S2GRAPH-73), [S2GRAPH-74](https://issues.apache.org/jira/browse/S2GRAPH-74) and [S2GRAPH-75](https://issues.apache.org/jira/browse/S2GRAPH-75). You can merge this pull request into a Git repository by running: $ git pull https://github.com/jongwook/incubator-s2graph feature/S2GRAPH-70 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/incubator-s2graph/pull/57.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #57 commit dcc19dd7caf34cc1c452959059dfcf19a5b78fe2 Author: Jong Wook Kim Date: 2016-05-17T01:06:15Z [S2GRAPH-70]: WIP - "sbt package" will gather all dependent libraries and bin/* conf/* to target/deploy/ commit b595bc9e8a5fc024c8a2bae714381bfe3f41 Author: Jong Wook Kim Date: 2016-05-23T23:50:01Z added bin/s2graph-daemon.sh that launches daemon processes. After doing 'activator clean package', the package for distribution will be prepared at ./target/deploy/, where you can execute commands like: bin/s2graph-daemon.sh start s2rest_play bin/s2graph-daemon.sh stop s2rest_play To start or stop a daemon process for s2rest_play. The following will run the server in the foreground bin/s2graph-daemon.sh run s2rest_play These files are internal scripts and will NOT be what end-users will be calling or modifying. For that purpose, `start-s2graph.sh`, `stop-s2graph.sh`, etc. is to be added. commit c1f9842f0dacee44f5e049d71c30ad3d97c4c518 Author: Jong Wook Kim Date: 2016-06-01T00:32:56Z [S2GRAPH-74] scripts to start and stop a standalone HBase server. Additionally, this commit includes: - A version of asynchbase that shades Google Guava, - Forcing Google Guava 12.0.1 to avoid their backward incompatibility with StopWatch commit 436ec0e62720fcf0827b4822379d6e186db917b1 Author: Jong Wook Kim Date: 2016-06-06T02:47:48Z [S2GRAPH-74] standalone HBase properly uses var/hbase commit 56e684784d520245d0cc31758bcab816cf318936 Author: Jong Wook Kim Date: 2016-06-06T22:10:47Z [S2GRAPH-75] Use H2 (in MySQL compatibility mode) as the default metastore, in the package created by 'activator package' > Automate the process of building a distribution package > --- > > Key: S2GRAPH-70 > URL: https://issues.apache.org/jira/browse/S2GRAPH-70 > Project: S2Graph > Issue Type: Improvement >Reporter: Jong Wook Kim > > Since the Vagrant image (which is the only documented way to trying out > S2Graph) is outdated, and also as a prerequisite of making a release, it is > desirable to be able to readily make a distribution package. The goal is to > be able for an end-user to download & untar the package, and run a shell > script to start off s2graph and all its dependencies. > AFAIK S2graph requires mysql for storing the metadata and HBase for the > actual data. Both could be contained in the package if: > - we run HBase server that is in the hbase-server jar, running without HDFS, > saving data in the local filesystem instead, and > - instead of the full MySQL server, use an embedded database like derby or h2 > to store the metadata. > It would also require some technical consideration about how exactly we > should layout the files and what toolchain we should use to build the > package, so I think adding child issues for each subtask would be appropriate. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
