how-to block ads
|
scrummie02
Bentley
Premium
join:2004-04-16
Arlington, VA
| CIsco 2950 slow network transfers
Hello all.
I have a 2950 catalyst 100MB switch. Connected to it I have my workstation, my WHS box. Linux box and a VMWare ESXi box. They're all in the same vlan. The other vlan is my IP phone, my vmware ESXi server (on another interface), and wireless.
In any case, on both vlans and with speeds on the NICS and the ports switched to the same (i.e. 100 MB full, auto negotiate), I'm getting very slow network transfer speeds with it comes to SMB transfers and NFS transfers. I'm only getting about 8 MBs. I am thinking I should be getting faster than that. I've been googleing but to no avail. Anyone run into this before?
--
'A government big enough to give you everything you want, is strong enough to take everything you have.'
Thomas Jefferson
| |
vipergg22
@netacc.net
| Its unclear how you have your interfaces configured. How are the switchports configured ? As auto ? If so then the nics on the servers "must" also be auto . You cannot set the server end as hardcoded and leave the switch as auto or vice versa . This creates a speed duplex mismatch if you do . If both ends match then you will have to look at server settings as the 2950 should be able to give the full 100 meg thruput on your ports . If you do a show interface status on your ports does it show 100/full ? If you do a show int count errors does it show any errors on those ports ? | |
scrummie02
Bentley
Premium
join:2004-04-16
Arlington, VA
| they're both set to auto and ther are some errors but not a lot. Here is the output from one interface:
Auto-duplex (Full), Auto Speed (100), 100BaseTX/FX
ARP type: ARPA, ARP Timeout 04:00:00
Last input 00:00:30, output 00:00:01, output hang never
Last clearing of "show interface" counters never
Queueing strategy: fifo
Output queue 0/40, 0 drops; input queue 0/75, 0 drops
5 minute input rate 6060000 bits/sec, 548 packets/sec
5 minute output rate 148000 bits/sec, 275 packets/sec
112048007 packets input, 3668108221 bytes
Received 263869 broadcasts, 0 runts, 0 giants, 0 throttles
8 input errors, 8 CRC, 0 frame, 11 overrun, 11 ignored
0 watchdog, 178699 multicast
0 input packets with dribble condition detected
91296795 packets output, 1851865801 bytes, 0 underruns
0 output errors, 0 collisions, 1 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out
FastEthernet0/2 is up, line protocol is up
Hardware is Fast Ethernet, address is 0002.fd74.0182 (bia 0002.fd74.0182)
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 15/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive not set
Auto-duplex (Full), Auto Speed (100), 100BaseTX/FX
ARP type: ARPA, ARP Timeout 04:00:00
Last input never, output 00:00:01, output hang never
Last clearing of "show interface" counters never
Queueing strategy: fifo
Output queue 0/40, 0 drops; input queue 0/75, 0 drops
5 minute input rate 148000 bits/sec, 275 packets/sec
5 minute output rate 6103000 bits/sec, 552 packets/sec
568557916 packets input, 1611993662 bytes
Received 2146056 broadcasts, 0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 1486061 multicast
0 input packets with dribble condition detected
342152395 packets output, 1074284809 bytes, 0 underruns
1853 output errors, 9072 collisions, 1 interface resets
0 babbles, 1853 late collision, 1007 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out
--
'A government big enough to give you everything you want, is strong enough to take everything you have.'
Thomas Jefferson
| |
Bink
join:2006-05-14
Denver, CO | reply to scrummie02
I think you might have the MB and Mb thing confused. On a Fast Ethernet/100Mbps switch, 8MBps is pretty good. | |
vipergg22
@netacc.net | reply to scrummie02
Also clear all the counters , you should not be seeing any collisions in a full duplex environment . If these continue then hardcode both the switchports and the nics to 100 full, there should be "0" collisions on the ports . | |
vipergg22
@netacc.net
| reply to scrummie02
That is possible , it is 100 megabits per seconds per port , if you are getting 8 megabytes transfer thats approx. 80 megabits per second which isn;t terrible . When you do a trAnsfer file look at the counters with a show interf. summary and see what is actually being transfered . You might want to turn the load interval down on the ports to like 60 seconds to get a better idea what the traffic, this will give a 1 minute rate versus the default of 5 minutes. | |
scrummie02
Bentley
Premium
join:2004-04-16
Arlington, VA | thanks! I'll try that. | |
NOCMan
Verizon Fios User
Premium
join:2004-09-30
Flower Mound, TX | Also check the duplex of the device connected to f0/2.
With Linux I always force both ends 100/Full..
mii-tool ethX --force 100BaseTX-FD
conf t
int f0/2
speed 100
duplex full.
See if the errors go away. | |
-
|
