我需要一些帮助来理解如何在REDIS中存储和索引以下JSON的所有元素。我的目标是在REDIS中为所有EC2实例存储所有AWS DESCRIBE_INSTACE CLI数据，并查询索引数据以返回具有特定匹配字符串的所有EC2实例。
对于EG - 我需要能够找到并返回属于特定SUBNET / VPC的所有EC2实例
可以在REDIS中完成吗？
"Reservations": [
{
"Instances": [
{
"Monitoring": {
"State": "disabled"
},
"PublicDnsName": "",
"State": {
"Code": 16,
"Name": "running"
},
"EbsOptimized": "false",
"LaunchTime": "xxxxxxxxx",
"PrivateIpAddress": "x.x.x.x",
"ProductCodes": [],
"VpcId": "xxxxx",
"StateTransitionReason": "",
"InstanceId": "i-xxxxxxx",
"EnaSupport": "true",
"ImageId": "ami-xxxxx",
"PrivateDnsName": "ip-xxxxxx.ec2.internal",
"KeyName": "xxxxxxv",
"SecurityGroups": [
{
"GroupName": "xxx",
"GroupId": "sg-xxxx"
},
{
"GroupName": "xxxxxx",
"GroupId": "sg-xxxxx"
},
{
"GroupName": "xxxxx",
"GroupId": "sg-xxxxxx"
},
{
"GroupName": "xxxxx",
"GroupId": "sg-xxxxxx"
}
],
"ClientToken": "xxxxx",
"SubnetId": "subnet-xxxxx",
"InstanceType": "t2.micro",
"NetworkInterfaces": [
{
"Status": "in-use",
"MacAddress": "xxxxxxxx",
"SourceDestCheck": "true",
"VpcId": "vpc-xxxxx",
"Description": "",
"NetworkInterfaceId": "eni-xxxxx",
"PrivateIpAddresses": [
{
"PrivateDnsName": "ip-xx-ec2.internal",
"Primary": "true",
"PrivateIpAddress": "xxxxx"
}
],
"PrivateDnsName": "ip-xxxx-xx.ec2.internal",
"Attachment": {
"Status": "attached",
"DeviceIndex": 0,
"DeleteOnTermination": "true",
"AttachmentId": "eni-attach-xxxxx",
"AttachTime": "2017-0xxxxx"
},
"Groups": [
{
"GroupName": "xx",
"GroupId": "sg-xxxx"
},
{
"GroupName": "xxxx",
"GroupId": "sg-xxx"
},
{
"GroupName": "xxxx",
"GroupId": "sg-xxx"
},
{
"GroupName": "xxxx",
"GroupId": "sg-xxxx"
}
],
"Ipv6Addresses": [],
"OwnerId": "xxx",
"SubnetId": "subnet-xxxx",
"PrivateIpAddress": "1xxxx"
}
],
"SourceDestCheck": "true",
"Placement": {
"Tenancy": "default",
"GroupName": "",
"AvailabilityZone": "us-xxxxxxx"
},
"Hypervisor": "xen",
"BlockDeviceMappings": [
{
"DeviceName": "/dev/xxxxxx",
"Ebs": {
"Status": "attached",
"DeleteOnTermination": "true",
"VolumeId": "vol-xxxxxx",
"AttachTime": "2017-xxxxxxx"
}
}
],
"Architecture": "x86_64",
"RootDeviceType": "ebs",
"IamInstanceProfile": {
"Id": "xxxxxxxx",
"Arn": "arn:aws:iam::xxxxxxx"
},
"RootDeviceName": "/dev/xxxxx",
"VirtualizationType": "hvm",
"Tags": [
{
"Value": "xxxxxx",
"Key": "aws:cloudformation:stack-name"
},
{
"Value": "xxxxxxx",
"Key": "aws:cloudformation:logical-id"
},
{
"Value": "arn:aws:cloudformation:xxxxxx",
"Key": "aws:cloudformation:stack-id"
}
],
"AmiLaunchIndex": 0
}
],
"ReservationId": "r-xxxxx",
"RequesterId": "xxxxx",
"Groups": [],
"OwnerId": "xxxxxx"
}
]
}
答案 0 :(得分：0)
是的，这是可以做到的。然而，重要的是要意识到 Redis 本身并不真正关心 JSON，因此您必须将数据转换为可以有效查询的格式。这种“规范化”（让我们暂时称之为）必须在应用程序级别发生。换句话说 - 你必须处理这个问题，而不是 Redis。
<块引用>对于 EG - 我需要能够找到并返回属于特定 SUBNET/VPC 的所有 EC2 实例
让我们看一下这个例子。对此数据建模的一种方法可能如下所示：
instance:{instance_id} 下的 JSON 编码 blob。instance_by_subnet:subnet-xxxxx 集下。更具体地说 - 使用您提供的示例 - 这可能如下所示：
SET instance:1 {"Monitoring":.......} # Store first instance
SADD instance_by_subnet:subnet-abc123 1 # Add instance id to relevant set
通过这种方式，您可以通过调用 SMEMBERS instance_by_subnet:subnet-abc123 来确定实例属于特定子集。您必须发出另一个 GET 命令才能实际读取 JSON blob。
有多种方法可以维护这些索引，还有许多其他数据结构在建模数据时会派上用场。如果您有兴趣，我写了 an article 更详细地介绍了这一点。