Dropped packets

In researching an issue with my backup solution, it was observed that my SLES 11 sp2 OES sp1 server running as a VM on VMWare 5.0 using the default network config, is dropping packets. The percentage of drop was over 2%. I removed the network adapter and rebooted the server, re-installed the adapter and the dropped rate is now about 1.5%. I also noticed that another VM running the same SLES config was performing about the same. Is this typical of a SLES box on VM? Is this a Linux problem or a VMWare problem? The version of SLES installed is not the VMWare version.

Any thoughts would be appreciated.

What do you mean by “dropped packets” exactly? How are you seeing them?

There have been some discussions in the last year or so about numbers now
showing up as “dropped” in the ‘ip’ command output (‘ip -s link’) but
those are due to a different way of counting packets. This does not mean
packets intended for SLES which it threw out just for fun, or at least
that has not been confirmed in the threads I’ve seen. Rather, the packets
are mangled already, where ‘mangled’ means either broken or at least not
something that the target SLES box is intended to receive (using a
protocol not intended for the SLES box to receive).

Without a bit more detail it is hard to say why the system may report
“dropped packets” but that percentage could be normal, especially compared
to older kernels, because more types of packets (previously just ignored)
are now being counted. The advice I’ve seen for those issues is
basically: “Stop sending nonsense packets.”


Good luck.

If you find this post helpful and are logged into the web interface,
show your appreciation and click on the star below…

On 23/01/2014 16:04, hydepark wrote:
[color=blue]

In researching an issue with my backup solution, it was observed that my
SLES 11 sp2 OES sp1 server running as a VM on VMWare 5.0 using the
default network config, is dropping packets. The percentage of drop was
over 2%. I removed the network adapter and rebooted the server,
re-installed the adapter and the dropped rate is now about 1.5%. I also
noticed that another VM running the same SLES config was performing
about the same. Is this typical of a SLES box on VM? Is this a Linux
problem or a VMWare problem? The version of SLES installed is not the
VMWare version.

Any thoughts would be appreciated.[/color]

By “VMware 5.0” do you mean ESXi 5.0 or are you talking about a desktop
product (Workstation, Player, Fusion, etc.)?

How up-to-date are your VMs and VMware?

When you say you removed then re-installed the network adapter do you
mean a virtual NIC your VM uses or a physical one in the VMware server
itself?

Have you installed the VMware Tools in your VM?

Which NIC driver are you using?

HTH.

Simon
SUSE Knowledge Partner


If you find this post helpful and are logged into the web interface,
please show your appreciation and click on the star below. Thanks.