Open AviralSri opened 7 years ago
Just comma separate them
index: topbeat,packetbeat
Hi,
I already tried with as you replied. My rule file is as follows:-
es_port: 9200
index: topbeat-*,packetbeat-*
smtp_host: ismtp.corp.company.com
es_host: xx.xx.xx.xx
realert:
minutes: 5
alert: email
timeframe:
minutes: 30
type: frequency
from_addr: abc@email.com
num_events: 1
email: ["kpr@email.com"]
name: multiple_indices
filter:
- range:
fs.used_p:
from: 0.50
to: 1
- term:
beat.hostname: server1
- term:
http.code: 200
- term:
beat.hostname: server2
My question is how the elastalert will know fs.used_p and beat.hostname: server1 will be queried against topbeat- AND http.code:200 and beat.hostname: server2 will be queried against packetbeat-
How about
filter:
- or:
- and:
- range:
fs.used_p:
from: 0.50
to: 1
- term:
beat.hostname: server1
- and
- term:
http.code: 200
- term:
beat.hostname: server2
Hi,
filter: - or: - and: - range: fs.used_p: from: 0.50 - term: beat.hostname: server1 - term: fs.device_name: D:\ - and: - term: http.code: 200 - term: beat.hostname: server2
I have tried the filer but when I get emails it is containing information only about packetbeat-* index , where server2 is having http.code more than 1 in last 30 minutes. No mention of server1 where fs.used_p is 0.91 for fs.device_name: D:. I have cross checked it in kibana.
Can you provide me a link for writing filters and using these and, not and or in more complicated scenarios.
Have you tried removing packetbeat and see if anything matches the other index? You can either use something like elastalert-test-rule --days=7 file.yaml
or elastalert --verbose --start 2016-12-08
to query over an entire week at a time just to see if you get any matches.
I am using elasticsearch, topbeat and packetbeat. I need to query topbeat- and packetbeat- in the same rule file. How the filters will be separated by index?? OR The filters will be queried against both indexes??