Closed Zokol closed 2 weeks ago
I noticed that arkime-container has no ports open for the Hedgedog to access. Could this be the reason for it not to be able to upload the PCAPs?
$ docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
b0a82dd0bcca ghcr.io/idaholab/malcolm/nginx-proxy:24.05.0 "/sbin/tini -- /usr/…" 5 days ago Up 2 hours (healthy) 0.0.0.0:443->443/tcp, 0.0.0.0:9200->9200/tcp malcolm-nginx-proxy-1
f2009539be92 ghcr.io/idaholab/malcolm/dashboards:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 5601/tcp malcolm-dashboards-1
9b25c1a079e4 ghcr.io/idaholab/malcolm/netbox:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 9001/tcp malcolm-netbox-1
ef90e1f3675f ghcr.io/idaholab/malcolm/pcap-monitor:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 30441/tcp malcolm-pcap-monitor-1
b3339fcea600 ghcr.io/idaholab/malcolm/logstash-oss:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 9001/tcp, 0.0.0.0:5044->5044/tcp, 9600/tcp malcolm-logstash-1
fe6966bf60eb ghcr.io/idaholab/malcolm/zeek:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) malcolm-zeek-1
ca4cca21f17a ghcr.io/idaholab/malcolm/arkime:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 8000/tcp, 8005/tcp, 8081/tcp malcolm-arkime-1
adba28348adb ghcr.io/idaholab/malcolm/dashboards-helper:24.05.0 "/sbin/tini -- /usr/…" 5 days ago Up 2 hours (healthy) 28991/tcp malcolm-dashboards-helper-1
5efd029a2321 ghcr.io/idaholab/malcolm/opensearch:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 9200/tcp, 9300/tcp, 9600/tcp, 9650/tcp malcolm-opensearch-1
a001cec5f055 ghcr.io/idaholab/malcolm/zeek:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours malcolm-zeek-live-1
97971de84bf1 ghcr.io/idaholab/malcolm/filebeat-oss:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 0.0.0.0:5045->5045/tcp malcolm-filebeat-1
b6b25026ad9a ghcr.io/idaholab/malcolm/suricata:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours malcolm-suricata-live-1
92aa4ffe010f ghcr.io/idaholab/malcolm/file-monitor:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 3310/tcp, 8440/tcp malcolm-file-monitor-1
a7056ab282ec ghcr.io/idaholab/malcolm/postgresql:24.05.0 "/sbin/tini -- /usr/…" 5 days ago Up 2 hours (healthy) 5432/tcp malcolm-netbox-postgres-1
3101899d6255 ghcr.io/idaholab/malcolm/redis:24.05.0 "/sbin/tini -- /usr/…" 5 days ago Up 2 hours (healthy) 6379/tcp malcolm-netbox-redis-1
97c490466a96 ghcr.io/idaholab/malcolm/redis:24.05.0 "/sbin/tini -- /usr/…" 5 days ago Up 2 hours (healthy) 6379/tcp malcolm-netbox-redis-cache-1
b2aab0708bb6 ghcr.io/idaholab/malcolm/freq:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 10004/tcp malcolm-freq-1
77b40795fd95 ghcr.io/idaholab/malcolm/pcap-capture:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours malcolm-pcap-capture-1
97bf3b101e65 ghcr.io/idaholab/malcolm/htadmin:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 80/tcp malcolm-htadmin-1
3bb803dea3c8 ghcr.io/idaholab/malcolm/file-upload:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 80/tcp, 127.0.0.1:8022->22/tcp malcolm-upload-1
f86b975aa52d ghcr.io/idaholab/malcolm/api:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) 5000/tcp malcolm-api-1
56214a1237a4 ghcr.io/idaholab/malcolm/arkime:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours malcolm-arkime-live-1
f07d519e74a5 ghcr.io/idaholab/malcolm/suricata:24.05.0 "/usr/bin/tini -- /u…" 5 days ago Up 2 hours (healthy) malcolm-suricata-1
No, as arkime-capture on hedgehog writes directly to opensearch (port 9200).
Hedgehog is not just globally broken, as we have a lot of regression and acceptance testing across a variety of platforms before release. So we just need to find out what's wrong with your system/configuration.
Checking /opt/sensor/sensor_ctl/logs as you mentioned in #451 is a good idea. You should be able to see the arkime capture logs and filebeat logs there, if they're complaining about the connection you'd see it there.
Also, looking at the logs on Malcolm (./scripts/logs
) there may be some indications there of what's going on (authentication issues, maybe)?
It seems that I was fooled by Arkime's default 1 hour window. The actual issue was that Hedgehog is three hours behind Malcolm's clock, so the data it sent was not visible until I changed the timeline filter.
However, I see no reason why Hedgehog is actually running different time, as I configured it to fetch time from Malcolm's IP in setup phase. This still might be something that needs some looking into.
Just to document the steps I took investigating the issue, in case it would be helpful for someone in the future:
netstat
/opt/sensor/sensor-ctl/log
contains connections to Malcolm's port 9200 with word bulk
and code 200
, which would indicate that the captured data is successfully sent to opensearchstats
-tab in Malcolm's Arkime web UI and saw that the node has actually sent several packets to Malcolm, from which I started to suspect that the issue might actually reside on time difference between the machinesHmm, okay, I somewhat recently (within the last 6 months or so) changed Hedgehog's code so that it should sync immediately upon setting up time synchronization from Malcolm using htpdate, but I wonder if you already had services running when it happened if it woudn't take effect until they've been restarted. I'll double-check that. Thanks for the info.
Describe the bug I see PCAPs in Hedgedog local disk Hedgedog can see Malcolm No arkime sessions or traffic is visible in Arkime
To Reproduce Steps to reproduce the behavior:
/home/sensor/net_cap
Expected behavior I would expect to see the contents of the pcap-files in Hedgedog to be visible in Malcolm/Arkime dashboard after short while.
Malcolm Version:
How are you running Malcolm? ISO installed (Malcolm on VM, Hedgedog on Raspberry Pi 4 8GB)