Logstash处理json格式日志文件的三种方法

假设日志文件中的每一行记录格式为json的，如：

{"Method":"JSAPI.JSTicket","Message":"JSTicket:kgt8ON7yVITDhtdwci0qeZg4L-Dj1O5WF42Nog47n_0aGF4WPJDIF2UA9MeS8GzLe6MPjyp2WlzvsL0nlvkohw","CreateTime":"2015/10/13 9:39:59","AppGUID":"cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d","_PartitionKey":"cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d","_RowKey":"1444700398710_ad4d33ce-a9d9-4d11-932e-e2ccebdb726c","_UnixTS":1444700398710}

默认配置下，logstash处理插入进elasticsearch后，查到的结果是这样的：

{
    "_index": "logstash-2015.10.16",
    "_type": "voip_feedback",
    "_id": "sheE9eXiQASMDVtRJ0EYcg",
    "_version": 1,
    "found": true,
    "_source": { "message": "{\"Method\":\"JSAPI.JSTicket\",\"Message\":\"JSTicket:kgt8ON7yVITDhtdwci0qeZg4L-Dj1O5WF42Nog47n_0aGF4WPJDIF2UA9MeS8GzLe6MPjyp2WlzvsL0nlvkohw\",\"CreateTime\":\"2015/10/13 9:39:59\",\"AppGUID\":\"cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d\",\"_PartitionKey\":\"cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d\",\"_RowKey\":\"1444700398710_ad4d33ce-a9d9-4d11-932e-e2ccebdb726c\",\"_UnixTS\":1444700398710}", "@version": "1", "@timestamp": "2015-10-16T00:39:51.252Z", "type": "voip_feedback", "host": "ipphone", "path": "/usr1/data/voip_feedback.txt" } }

即会将json记录做为一个字符串放到”message”下，但是我是想让logstash自动解析json记录，将各字段放入elasticsearch中。有三种配置方式可以实现。

第一种，直接设置format => json

    file {
        type => "voip_feedback" path => ["/usr1/data/voip_feedback.txt"] format => json sincedb_path => "/home/jfy/soft/logstash-1.4.2/voip_feedback.access" }

这种方式查询出的结果是：

{
    "_index": "logstash-2015.10.16",
    "_type": "voip_feedback",
    "_id": "NrNX8HrxSzCvLl4ilKeyCQ",
    "_version": 1,
    "found": true,
    "_source": { "Method": "JSAPI.JSTicket", "Message": "JSTicket:kgt8ON7yVITDhtdwci0qeZg4L-Dj1O5WF42Nog47n_0aGF4WPJDIF2UA9MeS8GzLe6MPjyp2WlzvsL0nlvkohw", "CreateTime": "2015/10/13 9:39:59", "AppGUID": "cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d", "_PartitionKey": "cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d", "_RowKey": "1444700398710_ad4d33ce-a9d9-4d11-932e-e2ccebdb726c", "_UnixTS": 1444700398710, "@version": "1", "@timestamp": "2015-10-16T00:16:11.455Z", "type": "voip_feedback", "host": "ipphone", "path": "/usr1/data/voip_feedback.txt" } }

可以看到，json记录已经被直接解析成各字段放入到了_source中，但是原始记录内容没有被保存

第二种，使用codec => json

    file {
        type => "voip_feedback" path => ["/usr1/data/voip_feedback.txt"] sincedb_path => "/home/jfy/soft/logstash-1.4.2/voip_feedback.access" codec => json { charset => "UTF-8" } }

这种方式查询出的结果与第一种一样，字段被解析，原始记录内容也没有保存

第三种，使用filter json

filter {
    if [type] == "voip_feedback" { json { source => "message" #target => "doc" #remove_field => ["message"] } } }

这种方式查询出的结果是这样的：

{
    "_index": "logstash-2015.10.16",
    "_type": "voip_feedback",
    "_id": "CUtesLCETAqhX73NKXZfug",
    "_version": 1,
    "found": true,
    "_source": { "message": "{\"Method222\":\"JSAPI.JSTicket\",\"Message\":\"JSTicket:kgt8ON7yVITDhtdwci0qeZg4L-Dj1O5WF42Nog47n_0aGF4WPJDIF2UA9MeS8GzLe6MPjyp2WlzvsL0nlvkohw\",\"CreateTime\":\"2015/10/13 9:39:59\",\"AppGUID\":\"cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d\",\"_PartitionKey\":\"cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d\",\"_RowKey\":\"1444700398710_ad4d33ce-a9d9-4d11-932e-e2ccebdb726c\",\"_UnixTS\":1444700398710}", "@version": "1", "@timestamp": "2015-10-16T00:28:20.018Z", "type": "voip_feedback", "host": "ipphone", "path": "/usr1/data/voip_feedback.txt", "Method222": "JSAPI.JSTicket", "Message": "JSTicket:kgt8ON7yVITDhtdwci0qeZg4L-Dj1O5WF42Nog47n_0aGF4WPJDIF2UA9MeS8GzLe6MPjyp2WlzvsL0nlvkohw", "CreateTime": "2015/10/13 9:39:59", "AppGUID": "cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d", "_PartitionKey": "cb54ba2d-1d38-45f2-9ed1-abff0bf7dd3d", "_RowKey": "1444700398710_ad4d33ce-a9d9-4d11-932e-e2ccebdb726c", "_UnixTS": 1444700398710, "tags": [ "111", "222" ] } }

可以看到，原始记录被保存，同时字段也被解析保存。如果确认不需要保存原始记录内容，可以加设置：remove_field => [“message”]

比较以上三种方法，最方便直接的就是在file中设置format => json

另外需要注意的是，logstash会在向es插入数据时默认会在_source下增加type,host,path三个字段，如果json内容中本身也含有type,host,path字段，那么解析后将覆盖掉logstash默认的这三个字段，尤其是type字段，这个同时也是做为index/type用的，覆盖掉后，插入进es中的index/type就是json数据记录中的内容，将不再是logstash config中配置的type值。

这时需要设置filter.json.target，设置该字段后json原始内容将不会放在_source下，而是放到设置的”doc”下：

{
    "_index": "logstash-2015.10.20",
    "_type": "3alogic_log",
    "_id": "xfj3ngd5S3iH2YABjyU6EA",
    "_version": 1,
    "found": true,
    "_source": { "@version": "1", "@timestamp": "2015-10-20T11:36:24.503Z", "type": "3alogic_log", "host": "server114", "path": "/usr1/app/log/mysql_3alogic_log.log", "doc": { "id": 633796, "identity": "13413602120", "type": "EAP_TYPE_PEAP", "apmac": "88-25-93-4E-1F-96", "usermac": "00-65-E0-31-62-5D", "time": "20151020-193624", "apmaccompany": "TP-LINK TECHNOLOGIES CO.,LTD", "usermaccompany": "" } } }

这样就不会覆盖掉_source下的type,host,path值
而且在kibana中显示时字段名称为doc.type,doc.id…