cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
4389
Views
3
Helpful
4
Replies
singhaam007
Participant

slow network performance

Hi there,

We are having some slow network performance issues when some open any document from network drive.

I tried it on my pc with wire shark and whenever I do feet slow, opening files I get there messages in wire shark

Server                my pc

Pc sends the request to server (SMB)

x.x.101.61         x.x.108.207     TCP 1514 [TCP segment of a reassembled PDU]

x.x.101.61         x.x.108.207     TCP 1514 [TCP segment of a reassembled PDU]

x.x.101.61         x.x.108.207     TCP 1514 [TCP segment of a reassembled PDU]

x.x.101.61         x.x.108.207     TCP 1514 [TCP segment of a reassembled PDU]

x.x.101.61         x.x.108.207     TCP 1514 [TCP segment of a reassembled PDU]

x.x.101.61         x.x.108.207     TCP 1514 [TCP segment of a reassembled PDU]

Server response in next packet (SMB)

These packets keeps repeating them self and I can see it is taking long time to open one file.

Is this normal behavior or any issue?

thanks

4 REPLIES 4
RKB04071984_2
Beginner

Hello,

Did you check the switches? Aren't there any interface error?
Sometimes the problem can be physical. A cable or a interface issue.

If you have checked it before I would say to look for errors in the server.

In a case of the problem in the switch the whole network would be facing this problem.

Please, if you have more details tell us.

Thanks

Ronaldo

Alessio Andreoli
Contributor

Hi Amrinder,

in this fragment of your capture there is nothing wrong. you should post more about your capture and performing the next check:

a) broadcast packets amount

b) retransmissions

c) mtu and mss values

The point b) is the most important from a network performance perspective but the point number c) is the only idea that your fragment could give. Possibly you have too much fragmentations in your network due to too big packets that need to be reassembled by TCP while kept in its window. TCP windowing can cause network service deterioration if it is not properly sized. To set MTU and MSS values can help fixing it.

Alessio

Hi Alessio and Ronaldo,

here is the netstat from my pc

IPv4 Statistics

  Packets Received                   = 176688
  Received Header Errors             = 0
  Received Address Errors            = 112
  Datagrams Forwarded                = 0
  Unknown Protocols Received         = 0
  Received Packets Discarded         = 172
  Received Packets Delivered         = 176451
  Output Requests                    = 121558
  Routing Discards                   = 0
  Discarded Output Packets           = 0
  Output Packet No Route             = 0
  Reassembly Required                = 126
  Reassembly Successful              = 63
  Reassembly Failures                = 0
  Datagrams Successfully Fragmented  = 63
  Datagrams Failing Fragmentation    = 0
  Fragments Created                  = 126

ICMPv4 Statistics

                            Received    Sent
  Messages                  149         149
  Errors                    0           0
  Destination Unreachable   0           0
  Time Exceeded             0           0
  Parameter Problems        0           0
  Source Quenches           0           0
  Redirects                 0           0
  Echos                     0           149
  Echo Replies              149         0
  Timestamps                0           0
  Timestamp Replies         0           0
  Address Masks             0           0
  Address Mask Replies      0           0

TCP Statistics for IPv4

  Active Opens                        = 1285
  Passive Opens                       = 22
  Failed Connection Attempts          = 4
  Reset Connections                   = 335
  Current Connections                 = 20
  Segments Received                   = 157619
  Segments Sent                       = 103809
  Segments Retransmitted              = 41

UDP Statistics for IPv4

  Datagrams Received    = 18570
  No Ports              = 265
  Receive Errors        = 0
  Datagrams Sent        = 17502

thanks

Hi there,

Can you tell me more about how you sized the network? Is it a layer 2 network or what? By the way I confirm you need to revise the all three points above indicated and one more actually very worrying ... Why do you have so many active sessions and so many TCP resets? Start capturing traffic for a hour or so and then analyse what suggested above. A netstat is just a snapshot of your data flows and can't truly help.

Rate if helps

Alessio

Sent from Cisco Technical Support iPad App