如何设置neo4j以使用Google Compute Engine？

Question

我想知道如何让neo4j与Google Compute Engine合作.有人这样做过吗？你遇到了什么问题？

Answer 1

干得好,

基本设置

• 安装和设置 gcloud
• 安装py2neo
• 使用图像创建您的GCE实例(https://console.developers.google.com/project/PROJECT_APPID/compute/instancesAdd)(debian-7-wheezy-v20141021,D​​ebian GNU/Linux 7.7(wheezy)amd64建立于2014-10 -21或任何)
• SSH你的实例gcloud compute ssh INSTANCE_NAME --zone AVAILABLE_ZONES- > AVAILABLE_ZONES
• neo4j在GCE中下载并安装- 您可能需要安装java(Fix)和lsof(Fix :) apt-get install lsof.

GCE的配置

• 配置服务器neo4j
• (可选),添加neo4j https支持
• Whitelist neo4j端口7474(有关网络和防火墙的更多信息)

• 从github添加安全用户名:密码

  gcloud compute firewall-rules create neo4j --network default --allow tcp:7474

玩转

• 启动neo4j服务器 ./bin/neo4j start
• 检查正在运行的实例@ http://IP_ADDRESS:7474/

一旦py2neo安装并启动服务器,请尝试一些pycode来测试它

>> from py2neo.neo4j import GraphDatabaseService, CypherQuery
>> # Set up a link to the local graph database.
>> # When () left blank defaults to http://localhost:7474/db/data/
>> graph = GraphDatabaseService('http://IP_ADDRESS:7474/db/data/')
>> CypherQuery(graph, "CREATE (n {name:'Example'}) RETURN n;").execute()

在python设置/代码之上,您也可以在GAE中使用它.

参考

• 检查GCE的价格.
• py2neo食谱
• 计算实例
• gcloud计算

编辑:Appengine + Neo4j

from py2neo import neo4j
GRAPH_DB = neo4j.GraphDatabaseService(
        'http://uname:psswd@localhost:7474/db/data/')
if IS_PROD:
    GRAPH_DB = neo4j.GraphDatabaseService(
        'http://uname:psswd@host:port/db/data/')

def _execute(query):
    """Execute all neo4j queries and return list of Record objects.

    Returns:
      Returns list of Record objects.
    """
    try:
        result = neo4j.CypherQuery(GRAPH_DB, query).execute()
        # logging.info(result.data)
        return result
    except neo4j.CypherError as error:
        logging.error(error.exception)
    except DeadlineExceededError as dead:
        logging.warn(dead)
    except urlfetch_errors.InternalTransientError as tra_error:
        logging.warn(tra_error)
    except httplib.HTTPException as exp:
        logging.warn(exp)
    except neo4j.http.SocketError as soc:
        logging.warn(soc)