Closed ZoccoCss closed 2 months ago
Hey @ZoccoCss it sounds like you are looking to configure alerting for your suricata logs. I understand that you are ingesting suricata logs via the wazuh agent but have dedicated pipeline and index for your suricata events. Dedicating a seperate pipeline and index for suricata logs is a great approach and is what I would have also recommened.
You are correct when configuring a Source Configuration of the Incident Management to select a suricata-_0
index. CoPilot, is then going to collect the Source
automatically by reading the syslog_type
field. This is further explained in our recent video at the 13:09 timestamp: https://youtu.be/3p6qiH9UF8U?si=fJ57kFRbGKFW7tDr&t=789.
Since you are ingesting suricata events via the wazuh agent, it is likely that the syslog_type
field is set to wazuh
. What you need to do is to change the syslog_type
to suricata
. This can most easily be done by adding the below pipeline rule to your Suricata pipeline:
rule "SYSLOG TYPE SURICATA"
when
$message.rule_group1 == "suricata"
then
set_field("syslog_type", "suricata");
end
Then Rotate active write index
for your Suricata index:
And select the new index name within CoPilot. Likely suricata-_1
which will then contain the syslog_type
of suricata
.
Hope that helps!
Thank you,
That solved the source issue. I also needed to change the alert_severity for the Alerts to work.
rule "SURICATA"
when
$message.rule_group1 == "suricata"
then
// adapt the fields
set_field("syslog_type", "suricata");
set_field("alert_severity", $message.data_alert_severity);
end
Now I'm facing the following two other issues.
1 - the alerts grouping seem to group alerts that are different. In the following screenshot you can see the two alerts grouped together.
2 - the alerts don't appear in the Incident Management Alerts. I believe it's because these alerts don't have a agent_labels_customer since OPNsense is freeBSD and this OS doesn't have a Wazuh group created in the provisioning.
So, I'm left with one of three choices.
1- Hardcode a agent_labels_customer code in the Suricata pipeline 2- Use the Linux Wazuh group for the freeBSD OPNsense agent 3- Create a new Wazuh group for freeBSD similar to the Linux group.
Please let me know which one is the most appropriate or if it's a different one.
Hey @ZoccoCss , I recommend hardcoding the agent_labels_customer
code within the Suricata Pipeline...make sure that this customer code reflects to the one that exists for that Customer within CoPilot.
Thank you.
Hardcoding agent_labels_customer did the trick since I only have one customer with the OPNsense router. If there were more I guess I would just need to make rules to separate them.
Describe the bug I'm trying to use Copilot to process the Suricata logs from an OPNsense firewall. The logs are being sent through the Wazuh agent and I use a stream and a Pipeline to separate the logs and be picked up by Copilot's SURICATA ALERT SEVERITY 1.
The alerts appear in the Alert section of Copilot, however, when I try to create a source for Incident Management for the Suricata index, I get a message that "A configuration for "wazuh" already exists. Please select a different Index name to proceed. ".
I've already created a source as described in your Case Management video and it uses the Wazuh as a source.
To Reproduce
Expected behavior I expected the creation of a new source to process the cases generated from Suricata Alerts
Screenshots
Side Question
I noticed that when selecting the source in the Source Configuration of the Incident Management, you have to select a specific index and you're not given the option of using a wildcard.
Therefore, do we need to constantly update the sources as Wazuh creates new indices?