AIX systems transmit data to the Suse system slow

Suse between AIX systems and data transmission system is very slow, WAN transmission by only 500k, but the data transfer rate between the Suse system is normal, you can reach 2M.

You may want to post in an AIX forum, since the problem only appears to be
happening there.

Out of curiosity, what kind of transmission? Has it ever been better?
Can you test with more than one AIX box? Do any other types of
transmissions to/from the AIX box go more-quickly (other protocols, other
ports, other machines)? Does it matter if you’re primarily sending to or
receiving from the AIX box? What if you remove the WAN link factor since
even 2 M(bit?) is pretty slow by any current LAN (vs. WAN) networking

Good luck.

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

Hi andyleigood,

we’d need much more details to get into the picture and to analyze the root cause.

As you’re not new here, I’ll assume that you mean “SLES” when you say “Suse” - am I right? If yes, which SLES?

Additionally, please tell us how you’re transmitting data, in terms of which protocol you use, which clients are involved, and which servers (piece of software).

Next, a short description of the environment might be helpful - a WAN link is involved, is this a physical dedicated connection terminating in routers in your environments? Some tunnel across the Internet? The server end of the connections seems to be a SLES server. The client end seems to be one or multiple AIX boxes, and one or more SLES boxes. Are these all in the same network (subnet/physical network/…)?

There are several factors that influence throughput, not only bandwidth, but also network-added delay. The latter has severe impact on non-streaming transmission protocols, IOW when the client has to wait for some acknowledgment before transmitting the next chunk of data. Typically, transmission windows are used for this, where different clients many use different window sizes, negotiated with the server end. That all by itself might be an explanation.

Are you capable of providing network traces when it boils down to network-specific issues?

In other words, your initial description can only help as a starting point for analyzing this issue - please provide details :wink: