Hi,

We had a problem , that our cluster is every 10..14 days was complete unreachable because
of small ECB's (I belive it at least).
So we decided to change the NIC from BroadCom B57 to Intel Server Adapters.

Now the Servers are running almost 3 weeks fine.

But......

In the NRM I see the Available ECB's decreasing (about 5...10 percent per day).

I have stopped almost every process, but I don't find any solution for this.
NRM shows me:

Server with SP7 (a view ressources)

E1000.LAN 128
TCPIP.NLM 9
TOTAL TRACKED ECBS 137


Server with SP6 (many ressources)

E1000.LAN 128
TCPIP.NLM 2979
TOTAL TRACKED ECBS 3107


and the ECP Info:

Allocated ECBs5000510020000
Free ECBs19935100 39 %
Get ECB Requests 1035209782023
Get ECB Fails 0 0
Error Allocating More ECBs 0 0


Should I worry about that or are the ECB's increasing by itself when a threshold is reached ?

The decreasing occurs after 9 PM and before 7 AM, but I don't have much pleasure to spend a night
stearing on the screen to see what and when it happend.

So the problem is on a NW 6.5 SP6 and SP7 too.

Any ideas for further troubleshooting ?

Thanks

Uwe