Hi Iban,
 
I think it's necessary more information to be able to help you. Here they are:
- Redhat version: Which is 7.2, 7.3 or 7.4?
- Redhat kernel version: In the FAQ of GPFS has the recommended kernel levels
- Platform: Is it x86_64?
- Is there a reason for you stay in 4.2.3-6? Could you update to 4.2.3-9 or 5.0.1?
- How is the name resolution? Can you do test ping from one node to another and it's reverse?
- TCP/IP tuning: What is the TCP/IP parameters you are using? I have used for 7.4 the following:
[root@XXXX sysctl.d]# cat 99-ibmscale.conf
net.core.somaxconn = 10000
net.core.netdev_max_backlog = 250000
net.ipv4.ip_local_port_range = 2000 65535
net.ipv4.tcp_rfc1337 = 1
net.ipv4.tcp_max_tw_buckets = 1440000
net.ipv4.tcp_mtu_probing = 1
net.ipv4.tcp_window_scaling = 1
net.ipv4.tcp_low_latency = 1
net.ipv4.tcp_max_syn_backlog = 4096
net.ipv4.tcp_fin_timeout = 10
net.core.rmem_default = 4194304
net.core.rmem_max = 4194304
net.core.wmem_default = 4194304
net.core.wmem_max = 4194304
net.core.optmem_max = 4194304
net.ipv4.tcp_rmem=4096 87380 16777216
net.ipv4.tcp_wmem=4096 65536 16777216
vm.min_free_kbytes = 512000
kernel.panic_on_oops = 0
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
vm.swappiness = 0
vm.dirty_ratio = 10
 
Since we disabled ipv6, we had to rebuild the kernel image with the following command:
[root@XXXX ~]# dracut -f -v
 
- GPFS tuning parameters: Can you list them?
- Spectrum Scale status: Can you send the following outputs:
  mmgetstate -a -L
  mmlscluster
  mmhealth cluster show
  mmhealth cluster show --verbose
  mmhealth node eventlog
  mmlsnode -L -N waiters
 
 
Abraços / Regards / Saludos,

 

Anderson Nobre
AIX & Power Consultant
Master Certified IT Specialist
IBM Systems Hardware Client Technical Team – IBM Systems Lab Services

community_general_lab_services
 

Phone: 55-19-2132-4317
E-mail:
IBM
 
 
----- Original message -----
From: Iban Cabrillo <[email protected]>
Sent by: [email protected]
To: [email protected]
Cc:
Subject: [gpfsug-discuss] Thousands of CLOSE_WAIT connections
Date: Fri, Jun 15, 2018 9:12 AM
 
Dear,
 
  We have reinstall recently from gpfs 3.5 to SpectrumScale 4.2.3-6 version redhat 7.
 
  We are running two nsd servers and a a gui, there is no firewall on gpfs network, and selinux is disable, I have checked changing the manager and cluster manager node between server with the same result, server 01 always increase the CLOSE_WAIT :
 
   Node  Daemon node name       IP address     Admin node name        Designation
--------------------------------------------------------------------------------
   1   gpfs01.ifca.es         10.10.0.111    gpfs01.ifca.es         quorum-manager-perfmon
   2   gpfs02.ifca.es         10.10.0.112    gpfs02.ifca.es         quorum-manager-perfmon
   3   gpfsgui.ifca.es        10.10.0.60     gpfsgui.ifca.es        quorum-perfmon
.......
 
Installation and configuration works fine, but now we see that one of the servers do not close the mmfsd connections and this growing for ever while the othe nsd servers is always in the same range:
 
[root@gpfs01 ~]# netstat -putana | grep 1191 | wc -l
19701
[root@gpfs01 ~]# netstat -putana | grep 1191 | grep CLOSE_WAIT| wc -l
19528
....
[root@gpfs02 ~]# netstat -putana | grep 1191 | wc -l
215
[root@gpfs02 ~]# netstat -putana | grep 1191 | grep CLOSE_WAIT| wc -l
0
 
this is causing that gpfs01 do not answer to cluster commands
 
NSD are balance between server (same size):
[root@gpfs02 ~]# mmlsnsd

 File system   Disk name    NSD servers                                    
---------------------------------------------------------------------------
 gpfs          nsd1         gpfs01,gpfs02
 gpfs          nsd2         gpfs01,gpfs02
 gpfs          nsd3         gpfs02,gpfs01
 gpfs          nsd4         gpfs02,gpfs01
.....
 
proccess seems to be similar in both servers, only mmccr is running on server 1 and not in 2
 
gpfs01
#######
root      9169     1  0 feb07 ?        22:27:54 python /usr/lpp/mmfs/bin/mmsysmon.py
root     11533  6154  0 13:41 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/mmsdrquery sdrq_fs_info all
root     11713     1  0 13:41 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/mmccrmonitor 15
root     12367 11533  0 13:43 ?        00:00:00 /usr/lpp/mmfs/bin/mmccr vget mmRunningCommand
root     12641  6162  0 13:44 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/mmsdrquery sdrq_nsd_info sdrq_nsd_name:sdrq_fs_name:sdrq_storage_pool
root     12668 12641  0 13:44 ?        00:00:00 /usr/lpp/mmfs/bin/mmccr fget -c 835 mmsdrfs /var/mmfs/gen/mmsdrfs.12641
root     12950 11713  0 13:44 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/mmccrmonitor 15
root     12959  9169 13 13:44 ?        00:00:00 /usr/lpp/mmfs/bin/mmccr check -Y -e
root     12968  3150  0 13:45 pts/3    00:00:00 grep --color=auto mm
root     19620 26468 38 jun14 ?        11:28:36 /usr/lpp/mmfs/bin/mmfsd
root     19701     2  0 jun14 ?        00:00:00 [mmkproc]
root     19702     2  0 jun14 ?        00:00:00 [mmkproc]
root     19703     2  0 jun14 ?        00:00:00 [mmkproc]
root     26468     1  0 jun05 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/runmmfs
 
 
[root@gpfs02 ~]# ps -feA | grep mm
root      5074     1  0 feb07 ?        01:00:34 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/mmccrmonitor 15
root      5128 31456 28 jun14 ?        06:18:07 /usr/lpp/mmfs/bin/mmfsd
root      5255     2  0 jun14 ?        00:00:00 [mmkproc]
root      5256     2  0 jun14 ?        00:00:00 [mmkproc]
root      5257     2  0 jun14 ?        00:00:00 [mmkproc]
root     15196  5074  0 13:47 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/mmccrmonitor 15
root     15265 13117  0 13:47 pts/0    00:00:00 grep --color=auto mm
root     31456     1  0 jun05 ?        00:00:00 /usr/lpp/mmfs/bin/mmksh /usr/lpp/mmfs/bin/runmmfs
 
Any idea will be appreciated.
Regards, I
 
_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at spectrumscale.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss
 

_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at spectrumscale.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss

Reply via email to