Bro + PF_RING:ZC: workers crash

Hello,
I’ve been using Bro and PF_RING:ZC but workers keep crashing. Most recently worker-1 crashed ~9 hours after deploying.

zbalance_ipc is run as :
/zbalance_ipc -i zc:em2 -l /var/log/zbalance.log -c 97 -n 6 -m 1 -a -r 0:dummy0 -r 1:dummy1 -r 2:dummy2 -r 3:dummy3 -r 4:dummy4 -r 5:dummy5

node.cfg:
[manager]
type=manager
host=localhost
pin_cpus = 2

[logger]
type=logger
host=localhost
pin_cpus = 3

[proxy-1]
type=proxy
host=localhost
pin_cpus = 4

[worker-1]
type=worker
host=localhost
interface=dummy0
pin_cpus = 5

[worker-2]
type=worker
host=localhost
interface=dummy1
pin_cpus = 6

[worker-3]
type=worker
host=localhost
interface=dummy2
pin_cpus = 7

[worker-4]
type=worker
host=localhost
interface=dummy3
pin_cpus = 8

[worker-5]
type=worker
host=localhost
interface=dummy4
pin_cpus = 9

[worker-6]
type=worker
host=localhost
interface=dummy5
pin_cpus = 10

broctl diag output for worker-1:
[worker-1]

No core file found.

Bro 2.5
Linux 3.10.0-693.17.1.el7.x86_64

Bro plugins:
Bro::AF_Packet - Packet acquisition via AF_Packet (dynamic, version 1.3)
Bro::PF_RING - Packet acquisition via PF_RING (dynamic, version 1.0)

==== No reporter.log

==== stderr.log
listening on dummy0

1529351356.071495 processing suspended
1529351356.071495 processing continued
1529382027.863765 received termination signal
1529382027.863765 469839129 packets received on interface dummy0, 1095837 dropped
/mnt/localraid/bro/share/broctl/scripts/run-bro: line 107: 110538 Killed nohup ${pin_command} $pin_cpu “$mybro” “$@”

==== stdout.log
max memory size (kbytes, -m) unlimited
data seg size (kbytes, -d) unlimited
virtual memory (kbytes, -v) unlimited
core file size (blocks, -c) unlimited

==== .cmdline
-i dummy0 -U .status -p broctl -p broctl-live -p local -p worker-1 local.bro broctl base/frameworks/cluster local-worker.bro broctl/auto

==== .env_vars
PATH=/mnt/localraid/bro/bin:/mnt/localraid/bro/share/broctl/scripts:/sbin:/bin:/usr/sbin:/usr/bin
BROPATH=/mnt/localraid/bro/spool/installed-scripts-do-not-touch/site::/mnt/localraid/bro/spool/installed-scripts-do-not-touch/auto:/mnt/localraid/bro/share/bro:/mnt/localraid/bro/share/bro/policy:/mnt/localraid/bro/share/bro/site
CLUSTER_NODE=worker-1

==== .status
TERMINATING [net_finish]

==== No prof.log

==== No packet_filter.log

==== No loaded_scripts.log

Are there any ways to get more detailed logs to help understand why the worker crashed? I restarted Bro with misc/profiling turned on, but these seems more like resource consumption statistics and won’t help to debug crashes.

Thanks

Pedantic, but that message looks like it was killed, not crashed. For
crashes, unless you don't compile w/ debug symbols, a stack trace
would be generated. For this case, maybe check system logs for more
info (e.g. a guess would be OOM killer).

- Jon