?
Solved

high latency and packet loss in lan

Posted on 2011-09-10
9
Medium Priority
?
2,128 Views
Last Modified: 2012-05-12
hi,

im using 2 cisco 3560e Giga,  with 4 etherchannel uplinks with vpan trunks, configured with hsrp on of the switches is active for all vlan ( i have 4 vlans).

i have 3 esx servers connected on both equipments for the HA.

Since a few time i have some packets lost if i ping between 2 machines in the same vlan or different vlan, the latency is minimum 1 ms average 2 ms and maximum 34 ms for repeat count of 1000 packets

the ping with maximum size make too much losts !!!

so any one have a suggestion ?
0
Comment
Question by:jamill
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
  • 3
  • 3
  • 2
  • +1
9 Comments
 
LVL 22

Expert Comment

by:eeRoot
ID: 36517610
how are the ESX switch ports configured
0
 
LVL 57

Expert Comment

by:giltjr
ID: 36517710
Define "maximum size".

How many are lost?

How busy is the server(s) while you are doing this?
0
 

Author Comment

by:jamill
ID: 36521424
Hello eeRoot

the esxi switchport is configured like ine the scheme 1 cisco
0
Building an interactive eFuture classroom

Watch and learn how ATEN provided a total control system solution including seamless switching matrix switch, HDBaseT extenders, PDU, lighting control to build an interactive eFuture classroom.

 

Author Comment

by:jamill
ID: 36521476

hello giltjr

here is a sample of a ping

ping 10.10.10.52 repeat 1000 size 1800

Type escape sequence to abort.
Sending 1000, 1800-byte ICMP Echos to 10.10.10.52, timeout is 2 seconds:
!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!
Success rate is 99 percent (994/1000), round-trip min/avg/max = 1/8/151 ms
0
 
LVL 57

Expert Comment

by:giltjr
ID: 36524427
Well, although I would expect 100%, 6 dropped packets out of 1000 is not that bad.

What is your MTU?   Have you enabled jumbo frames?

You are using 1800 bytes.  The default LAN MTU is 1500 bytes.  If you have the default then you are fragmenting every packet.

If you want to test without fragmenting you need to specify 1472 instead of 1800, unless of course you have jumbo frame enabled.  If you have jumbo frames you need to specific "MTU-28" (IP and ICMP headers are 28 bytes together).

What happens if you do not specify a size and let it default?
0
 
LVL 22

Expert Comment

by:eeRoot
ID: 36526376
Can you show the port configs from the ports on the 3560's that are connected to VMNIC's 1 and 2?
0
 

Author Comment

by:jamill
ID: 36527671
hello giltjr

My MTU is the standard one 1500.
No im not using jumbo frames so should i ?
when i do not specify the size, i have 0% of loss and the maximum latency time is reducing



0
 
LVL 24

Expert Comment

by:rfc1180
ID: 36530616
Well, although I would expect 100%, 6 dropped packets out of 1000 is not that bad.
I would disagree, any amount of packet loss is bad, especially utilizing TCP. Any packet loss is considered to be the result of network congestion and the congestion window size is reduced dramatically as a precaution. Your packet loss could be a result of many different factors and without some detailed troubleshooting and information, we will not be able to assist, we would only be assuming. please provide the configuration of your 2 switches, ensure that your port descriptions on configured on the ports so that we can identify what NICs are connected to what ports. Also, provide some screen shots on the vswitch configuration.

just as a test to rule on multihoming issues, remove NIC 2, and retest with 1800 bytes and report back.

Billy
0
 
LVL 57

Accepted Solution

by:
giltjr earned 2000 total points
ID: 36531706
Try specifying a size of 1472 and see what happens.

As I stated before, going larger than what a single Ethernet frame (MTU) causes fragmentation.  Once you have fragmentation there are other issues that can cause lost packets.

0

Featured Post

Technology Partners: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Many network operators, engineers, and administrators do not take several factors into consideration when troubleshooting network throughput and latency issues.  They often  measure the throughput by performing a measurement  by transferring a large…
Load balancing is the method of dividing the total amount of work performed by one computer between two or more computers. Its aim is to get more work done in the same amount of time, ensuring that all the users get served faster.
NetCrunch network monitor is a highly extensive platform for network monitoring and alert generation. In this video you'll see a live demo of NetCrunch with most notable features explained in a walk-through manner. You'll also get to know the philos…
Michael from AdRem Software outlines event notifications and Automatic Corrective Actions in network monitoring. Automatic Corrective Actions are scripts, which can automatically run upon discovery of a certain undesirable condition in your network.…
Suggested Courses

752 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question