Hi, we have a moderately loaded box (10.3-RELEASE-p11) running ruby on rails servers inside jails. They work fine, but after some uptime measured in weeks/months we can no longer see the listen queues of some of them:
Where did they go? Also, there's lots of messages of this kind:
Thanks for any tips.
$ sudo netstat -Lan|fgrep 127.0.0.60 $ jls
Code:
...
JID IP Address Hostname Path
3 127.0.0.60 an.example.com /var/jails/an.example.com
...Where did they go? Also, there's lots of messages of this kind:
Code:
May 30 13:37:16 myhost kernel: sonewconn: pcb 0xfffff808e1d00ab8: Listen queue overflow: 1 already in queue awaiting acceptance (2 occurrences)
May 30 13:39:51 myhost kernel: sonewconn: pcb 0xfffff8072919f188: Listen queue overflow: 1 already in queue awaiting acceptance (4 occurrences)
May 30 13:41:25 myhost kernel: sonewconn: pcb 0xfffff8072919f188: Listen queue overflow: 1 already in queue awaiting acceptance (5 occurrences)
May 30 13:43:00 myhost kernel: sonewconn: pcb 0xfffff808e1d00ab8: Listen queue overflow: 1 already in queue awaiting acceptance (4 occurrences)
May 30 13:44:20 myhost kernel: sonewconn: pcb 0xfffff80fdb43e310: Listen queue overflow: 1 already in queue awaiting acceptance (2 occurrences)
May 30 13:45:58 myhost kernel: sonewconn: pcb 0xfffff80fdb43e310: Listen queue overflow: 1 already in queue awaiting acceptance (5 occurrences)
May 30 13:47:50 myhost kernel: sonewconn: pcb 0xfffff8072919f188: Listen queue overflow: 1 already in queue awaiting acceptance (1 occurrences)
May 30 13:48:50 myhost kernel: sonewconn: pcb 0xfffff8072919f188: Listen queue overflow: 1 already in queue awaiting acceptance (2 occurrences)
May 30 13:50:08 myhost kernel: sonewconn: pcb 0xfffff8072919f188: Listen queue overflow: 1 already in queue awaiting acceptance (3 occurrences)
Code:
$ cat /boot/loader.conf
# This configuration file is used by loader.efi, which is executed when booting in *BIOS* mode (only).
# DO NOT EDIT this file unless you know what you are doing !
# See the /boot/loader.rc file within the EFI partition when booting in UEFI mode.
kern.geom.label.gptid.enable="0"
zfs_load="YES"
net.fibs=4
carp_load="YES"
ahci_load="YES"
accf_http_load="YES"
accf_data_load="YES"
aio_load="YES"
kern.hz=250
kern.maxdsiz="2048M"
kern.dfldsiz="2048M"
kern.maxssiz="2048M"
kern.ipc.maxpipekva="2048M"
kern.ipc.semmni=512
kern.ipc.semmns=1024
kern.ipc.semmnu=512
kern.maxproc="12328"
kern.maxprocperuid="11094"
kern.ipc.shm_use_phys=1
kern.ipc.nmbclusters=131072
kern.ipc.maxsockbuf=524288
kern.ipc.nsfbufs=10240
kern.sync_on_panic=1
net.inet.tcp.tcbhashsize=16384
vfs.zfs.arc_max="32G"
Code:
net.link.ether.inet.log_arp_movements=0
net.inet6.ip6.accept_rtadv=0
net.inet6.ip6.auto_linklocal=0
net.inet.ip.fw.dyn_buckets=8096
net.inet.ip.fw.verbose_limit=1000
net.inet.ip.fw.verbose=1
net.inet.tcp.fast_finwait2_recycle=1
net.inet.tcp.finwait2_timeout=15000
net.inet.tcp.drop_synfin=1
net.inet.ip.fw.dyn_max=65536
kern.maxfiles=62020
kern.maxfilesperproc=22190
kern.ipc.somaxconn=4096
kern.ipc.somaxconn=8192
net.inet.ip.portrange.hifirst=10000
net.inet.ip.intr_queue_maxlen=5120
net.inet.tcp.ecn.enable=1
# Postgresql
kern.ipc.shmmax=4294967296
# kern.ipc.shmall = kern.ipc.shmmax / hw.pagesize
kern.ipc.shmall=1048576
# Misc
## To disable closed port RST responses
net.inet.tcp.blackhole=0
net.inet.udp.blackhole=0
net.inet.icmp.icmplim_output=0
## For readproctitle
kern.ps_arg_cache_limit=512
# UFS read-ahead http://ivoras.sharanet.org/blog/tree/2010-11-19.ufs-read-ahead.html
vfs.read_max=256Thanks for any tips.
