Hi, a bit more information.

I tried adding a custom analyzer based off a recommendation I saw online 
somewhere.  This partly works in that it's not tokenising.  But I can't do 
wildcard searches in Kibana on the fields, and they're now case sensitive :(

curl localhost:9200/_template/logstash-username -XPUT -d '{
    "template": "logstash-*",
    "settings" : {
        "analysis": {
            "analyzer": {
                "lc_analyzer": {
                    "type": "custom",
                    "tokenizer": "keyword",
                    "filters": ["lowercase"]
                }
            }
        }
    },
    "mappings": {
        "_default_": {
             "properties" : {
                "User_Name" : { "type" : "string", "analyzer" : 
"lc_analyzer" }
            }
        }
    }
}'

Thanks

On Wednesday, January 8, 2014 3:26:03 PM UTC, Chris H wrote:
>
> Hi.  I've deployed elasticsearch with logstash and kibana to take in 
> Windows logs from my OSSEC log server, following this guide: 
> http://vichargrave.com/ossec-log-management-with-elasticsearch/
> I've tweaked the logstash config to extract some specific fields from the 
> logs, such as User_Name.  I'm having some issues searching on these fields 
> though.
>
> These searches work as expected:
>
>    - User_Name: * 
>    - User_Name: john.smith
>    - User_Name: john.*
>    - NOT User_Name: john.*
>
> But I'm having problems with Computer accounts, which take the format 
> "w-dc-01$" - they're being split on the "-" and the "$" is ignored.  So a 
> search for "w-dc-01" returns all the servers named "w-<anything>".  Also I 
> can't do "NOT User_Name: *$" to exclude computer accounts.
>
> The mappings are created automatically by logstash, and GET 
> /logstash-2014.01.08/_mapping shows:
>
> "User_Name": {
>
>    "type": "multi_field",
>    "fields": {
>       "User_Name": {
>          "type": "string",
>          "omit_norms": true
>       },
>       "raw": {
>          "type": "string",
>          "index": "*not_analyzed*",
>          "omit_norms": true,
>          "index_options": "docs",
>          "include_in_all": false,
>          "ignore_above": 256
>       }
>    }
> },
>
> My (limited) understanding is that the "not_analyzed" should stop the 
> field being split, so that my searching matches the full name, but it 
> doesn't.  I'm trying both kibana and curl to get results.
>
> Hope this makes sense.  I really like the look of elasticsearch, but being 
> able to search on extracted fields like this is pretty key to me using it.
>
> Thanks.
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/elasticsearch/96e74e53-54f9-48ec-9e5c-8f1354b264be%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to