Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

INFO::

I have a 6 HP ProLiant BL460c G7 servers in my Blade Chassis. RHEL6.2 64 bit is installed. Every blade server NIC1 is in use and Chassis LAN cable connected with switch only. Switch is configured in AUTO negotiation Mode.

PROBLEM::

When the network load is medium over TCP, I get hundreds of RX errors in all servers. Out of six on two is production servers(active/standby Mode) and getting maximum traffic in Rx error.

I am running org website over the same production servers(non commercial project). Generally site would have <500 Concentrate Establish session And in peak hour it reaches to > 500 I observe while ESTABLISH session are <100 then no movement in RX Errors I am sharing detail of active production server

Here is Kernal Parameter set at my server to extend the TCP buffer values :

[root@ ~]# sysctl -p
net.ipv4.ip_forward = 0
net.ipv4.conf.default.rp_filter = 1
net.ipv4.conf.default.accept_source_route = 0
kernel.sysrq = 0
kernel.core_uses_pid = 1
net.ipv4.tcp_syncookies = 1
kernel.msgmnb = 65536
kernel.msgmax = 65536
kernel.shmmax = 68719476736
kernel.shmall = 4294967296
net.ipv4.ip_local_port_range = 10000 65500
fs.aio-max-nr = 1048576
fs.file-max = 100000
net.core.rmem_max = 16777216
net.core.wmem_max = 16777216
net.core.rmem_default = 16777216
net.core.wmem_default = 16777216
net.core.optmem_max = 40960
net.ipv4.tcp_rmem = 4096 87380 16777216
net.ipv4.tcp_wmem = 4096 65536 16777216
net.core.netdev_max_backlog = 50000
net.ipv4.tcp_max_syn_backlog = 32768
net.core.somaxconn = 2048
net.ipv4.tcp_max_tw_buckets = 2000000
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_fin_timeout = 20
net.ipv4.tcp_slow_start_after_idle = 0
net.ipv4.udp_rmem_min = 8192
net.ipv4.udp_wmem_min = 8192
net.ipv4.conf.all.send_redirects = 0

net.ipv4.tcp_sack = 1
net.ipv4.tcp_window_scaling = 1
net.ipv4.tcp_keepalive_time = 600
net.ipv4.tcp_keepalive_intvl = 60
net.ipv4.tcp_keepalive_probes = 3
SYSTEM Detail is here ::
[root@ ~]# uname -a
Linux mohit 2.6.32-220.el6.x86_64 #1 SMP Wed Nov 9 08:03:13 EST 2011 x86_64 x86_64 x86_64 GNU/Linux
[root@ ~]# netstat -i
Kernel Interface table
Iface       MTU Met    RX-OK RX-ERR RX-DRP RX-OVR    TX-OK TX-ERR TX-DRP TX-OVR Flg
eth0       1500   0 43061491   2394      0      0 39363458      0      0      0 BMRU
lo        16436   0 12792095      0      0      0 12792095      0      0      0 LRU
You have mail in /var/spool/mail/root
[root@ ~]# ethtool eth0
Settings for eth0:
        Supported ports: [ TP ]
        Supported link modes:   1000baseT/Full
                                10000baseT/Full
        Supports auto-negotiation: Yes
        Advertised link modes:  1000baseT/Full
                                10000baseT/Full
        Advertised pause frame use: No
        Advertised auto-negotiation: No
        Speed: 1000Mb/s
        Duplex: Full
        Port: Twisted Pair
        PHYAD: 0
        Transceiver: internal
        Auto-negotiation: on
        MDI-X: Unknown
        Supports Wake-on: g
        Wake-on: d
        Link detected: yes

[root@ ~]# ethtool -S eth0|grep error
     rx_errors: 2394
     tx_errors: 0
     rx_crc_errors: 0
     rx_alignment_symbol_errors: 0
     rx_in_range_errors: 0
     rx_out_range_errors: 0
     rx_address_match_errors: 6932
share|improve this question
    
RX-ERR are malformed frames ("buggy packets" says the man page). What is the packet rate (packets per second) during the times logged? –  mbrownnyc Feb 14 at 19:33
    
possible duplicate of dropped packets on the interface –  mbrownnyc Feb 14 at 19:34
    
Dear mbrownnyc, nearly > 500 or < 1000 packets per second . Thanks –  Mohit M Feb 14 at 20:04
    
In TCP Dump getting huge amount of duplicat Ack packets.... Are you meaning these such packests are actully Rx Error packets. –  Mohit M Feb 14 at 20:18
    
I mean nothing. If you are observing an error, you have to fully understand what you are observing. So, I tried to define what RX-ERR actually means (it's really very vague). You've got to make sure you understand what you are using to diagnose an error. That post I linked to should give you some guidance on generally troubleshooting the error. I'm not sure if dup ACKs would increment this counter. Do you see a direct correlation? –  mbrownnyc Feb 14 at 20:31

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.