p520 AIX 5.2 ML05 - slow speed with GIGA interface

[MoshiachNow]

HI,

The new machine with ML05 shows good speed with built in GIGA versus other GIGA connections.
However it gives a very poor speed (100KB/sec) when communicating to 100BT stations.
Any ideas?

lsattr -El ent0
alt_addr 0x000000000000 Alternate ethernet address True
busintr 119 Bus interrupt level False
busmem 0xf0120000 Bus memory address False
chksum_offload yes Enable hardware transmit and receive checksum True
compat_mode no Gigabit Backward compatability True
copy_bytes 2048 Copy packet if this many or less bytes True
failover disable Enable failover mode True
flow_ctrl yes Enable Transmit and Receive Flow Control True
intr_priority 3 Interrupt priority False
intr_rate 10000 Max rate of interrupts generated by adapter True
jumbo_frames no Transmit jumbo frames True
large_send yes Enable hardware TX TCP resegmentation True
media_speed Auto_Negotiation Media speed True
rom_mem 0xf00c0000 ROM memory address False
rx_hog 1000 Max rcv buffers processed per rcv interrupt True
rxbuf_pool_sz 2048 Rcv buffer pool, make 2X rxdesc_que_sz True
rxdesc_que_sz 1024 Rcv descriptor queue size True
slih_hog 10 Max Interrupt events processed per interrupt True
tx_que_sz 8192 Software transmit queue size True
txdesc_que_sz 512 TX descriptor queue size True
use_alt_addr no Enable alternate ethernet address True
=====================================
lscfg -vl ent0
ent0 U787A.001.DPM04VB-P1-T5 2-Port 10/100/1000 Base-TX PCI-X Adapter (14108902)

2-Port 10/100/1000 Base-TX PCI-X Adapter:
Network Address.............00096B1B4F7C
ROM Level (alterable).......DV0210
Device Specific.(YL)........U787A.001.DPM04VB-P1-T5
=======================================
no -a
arpqsize = 12
arpt_killc = 20
arptab_bsiz = 7
arptab_nb = 73
bcastping = 0
clean_partial_conns = 0
ddp_recvspace = 9376
ddp_sendspace = 586
ddpcksum = 0
ddpforwarding = 1
ddpraw_recvspace = 19168
ddpraw_sendspace = 599
delayack = 0
delayackports = {}
dgd_packets_lost = 3
dgd_ping_time = 5
dgd_retry_time = 5
directed_broadcast = 0
extendednetstats = 0
fasttimo = 200
icmp6_errmsg_rate = 10
icmpaddressmask = 0
ie5_old_multicast_mapping = 0
ifsize = 256
inet_stack_size = 16
ip6_defttl = 64
ip6_prune = 1
ip6forwarding = 0
ip6srcrouteforward = 1
ip_nfrag = 200
ipforwarding = 1
ipfragttl = 60
ipignoreredirects = 0
ipqmaxlen = 100
ipsendredirects = 1
ipsrcrouteforward = 1
ipsrcrouterecv = 0
ipsrcroutesend = 1
llsleep_timeout = 3
lo_perf = 1
lowthresh = 90
main_if6 = 0
main_site6 = 0
maxnip6q = 20
maxttl = 255
medthresh = 95
multi_homed = 1
nbc_limit = 891289
nbc_max_cache = 131072
nbc_min_cache = 1
nbc_ofile_hashsz = 12841
nbc_pseg = 0
nbc_pseg_limit = 1957888
ndp_mmaxtries = 3
ndp_umaxtries = 3
ndpqsize = 50
ndpt_down = 3
ndpt_keep = 120
ndpt_probe = 5
ndpt_reachable = 30
ndpt_retrans = 1
net_malloc_police = 0
nonlocsrcroute = 0
nstrpush = 8
passive_dgd = 0
pmtu_default_age = 10
pmtu_rediscover_interval = 30
psebufcalls = 20
psecache = 1
pseintrstack = 12288
psetimers = 20
rfc1122addrchk = 0
rfc1323 = 1
rfc2414 = 0
route_expire = 1
routerevalidate = 0
rto_high = 64
rto_length = 13
rto_limit = 7
rto_low = 1
sack = 0
sb_max = 1048576
send_file_duration = 300
site6_index = 0
sockthresh = 85
sodebug = 0
somaxconn = 1024
strctlsz = 1024
strmsgsz = 0
strthresh = 85
strturncnt = 15
subnetsarelocal = 1
tcp_bad_port_limit = 0
tcp_ecn = 0
tcp_ephemeral_high = 65535
tcp_ephemeral_low = 32768
tcp_finwait2 = 1200
tcp_init_window = 0
tcp_inpcb_hashtab_siz = 24499
tcp_keepcnt = 8
tcp_keepidle = 14400
tcp_keepinit = 150
tcp_keepintvl = 150
tcp_limited_transmit = 1
tcp_maxburst = 0
tcp_mssdflt = 1460
tcp_nagle_limit = 65535
tcp_ndebug = 100
tcp_newreno = 1
tcp_nodelayack = 0
tcp_pmtu_discover = 1
tcp_recvspace = 245760
tcp_sendspace = 245760
tcp_tcpsecure = 0
tcp_timewait = 1
tcp_ttl = 60
thewall = 1048576
udp_bad_port_limit = 0
udp_ephemeral_high = 65535
udp_ephemeral_low = 32768
udp_inpcb_hashtab_siz = 24499
udp_pmtu_discover = 1
udp_recvspace = 42080
udp_sendspace = 9216
udp_ttl = 30
udpcksum = 1
use_isno = 1
#


Long live king Moshiach !

http://www.7for70.com/

 

[RMGBELGIUM]

The fact that the card is set to autonegotiate can sometimes give strange phenomena.
can you post the output of :

netstat -v

greetz,

R.

 

[DukeSSD]

MoshiachNow,
thread52-948577 - what speed is it actually running ?
Can you be sure the network guys have forced the switch speed ?

 

[MoshiachNow]

HI,

All switch/card setting were OK.I had to have the AUTONEGOTIATION on to get the GIGA speed,this is the only option on this card.
However,I have found the adnswer on some google forums:

"IBM guys comments:

There have been some reported problems with FC 5700, 5701, 5706 and 5707 which are the newer Gigabit Ethernet adapters we support in the pSeries systems. The devices have two TCP offload features, large_send (TCP Resegmentation) and chksum_offload (Checksum Offload), that are intended
to be performance features. In some instances the devices work as designed, but there have been issues with some software combinations where the offload features have caused problems. It is not difficult to disable the two features, but you want to make sure you have them either on or off for all devices that used the feature. To disable the large_send and chksum_offload you will detach the IP interfaces, put the device into a defined state, change the device configuration then reinitialized the
device and its interfaces. Sometimes it is easier to reboot the server as their may be application issues when IP interfaces are temporarily removed. The example shows how to change one device, ent1, and if you have other devices of the same type you will want to modify them as well.

1. If you can reboot:

chdev -l ent1 -a large_send=no -a chksum_offload=no -P
shutdown -Fr

2. If you can't reboot, but can take a very short communications outage
Note: if you get a device busy error when running the rmdev command then
try the ifconfig sequences again. If
you continue to get device busy error then use proceedure 1.

#!/bin/ksh
ifconfig en0 down detach
ifconfig et0 down detach
rmdev -l ent0
chdev -l ent0 -a large_send=no -a chksum_offload=no
# Checks to make sure device driver is unloaded
genkex | grep /usr/lib/drivers/pci/goentdd
# If above command returned no output then the driver is unloaded. If not, use proceedure 1.
mkdev -l ent0
mkdev -l en0
mkdev -l inet0
exit
"

Long live king Moshiach !

http://www.7for70.com/

 

[DukeSSD]

Thanks MoshiachNow,
I'd heard there were jumbo_frames problems but didn't realise there were large_send problems too.