RE: [lwip-users] Problem with long term running of lwIP.

[Pettinato, Jim]

I've been fighting something that sounds at least similar - at particular customer sites - for over a month. For me, it's locking up about every 48-60 hours, with a casually polling Modbus TCP connection (~ 200 ms between polls) - and lwIP 1.2.0. It requires a power cycle of our equipment to restore the operation of the stack.

 

I have had this reported in only two locations out of many however... and have not been able to duplicate it in the lab.

---

From: address@hidden [mailto:address@hidden On Behalf Of Taranowski, Thomas (SWCOE)
Sent: Wednesday, February 27, 2008 5:31 PM
To: Mailing list for lwIP users
Subject: [lwip-users] Problem with long term running of lwIP.

I have an application in which is receiving and sending UDP traffic over a long-term period.  After about 6 hours, something flakes out.  I was hoping I could get some input on possible causes.  Any ideas here on what the problem could be?  Is anyone else having a similar problem?

 

Other data:

I’m sending and receiving 50 packets/s over the period of 6 hours, at a fairly steady rate. (~1,000,000 packets)

Some packets are fragmented, others are only about 100 bytes.

 

 

The symptoms:

Somewhat random

Sometimes ping stops working, other times it still works.

Sometimes one UDP port stops working, other times all of them cease to function.

 

I’m running a cvs head from around Dec 3.  At the time, the head was stable and looked pretty solid.