read returned error 22 (Inv alid argument)

[Yvonne Gaffney]

Receiving the error in the log below.  Has anyone else received this type of error?  We are currently trying to send via a T1 connection.  At this time I am showing we are connected and have several messages pending.  I also have this interface set up to wait indefinitely for an acknowledgment.   The receiving side says they are not receiving anything and my side shows I am trying.

ci@clovth1:/quovadx/qdx5.6/integrator/cerner_id11/exec/processes/UMC>  *.err < [prod:prod:WARN/0:      UMC_cmd:09/01/2010 09:17:23] Output log file cycled [pdl :PDL :ERR /0:im_umc_orders:09/01/2010 10:03:12] read returned error 22 (Inv alid argument) [pdl :PDL :ERR /0:im_umc_orders:09/01/2010 10:03:12] PDL signaled exception: cod e 1, msg device error (remote side probably shut down) [pdl :PDL :ERR /0:im_umc_orders:09/01/2010 10:05:58] read returned error 22 (Inv alid argument) [pdl :PDL :ERR /0:im_umc_orders:09/01/2010 10:05:58] PDL signaled exception: cod e 1, msg device error (remote side probably shut down)

Topic

[Russ Ross]

When the receiving side manually shuts down their listener does your outbound Cloverleaf sending interface go from UP to OPENING?

If it does then that is a good start to know your interfaces is pointed to their interface.

If that test outcome confirms this then I would do a TCP/IP trace on the IP and port of the receiving interface.

If you don’t know how to perform such a TCP/IP trace you could try turning EO config to all.

Personally I have been challenged to understand the OE all output and just use the TCP/IP trace to look under the covers to see what is really going on.

Once you see what is happening under the covers share your finding and perhaps others on this forum can make useful suggestions.

Also, when you say T1 line that makes me want to ask if you are doing an interface outsied you enterprise or thru a VPN or any firewalls?

Russ Ross
RussRoss318@gmail.com

[Yvonne Gaffney]

To answer your questions….1) Yes when I bounce the interface is does go to an opening session – we have discovered that they must come up first or we will not resolve the connection issue.  2) Our T1 connection is not using VPN but firewalls on both sides.

Here is my result of the trace route I did –>

ci@clovth1:/quovadx/qdx5.6/integrator/cerner_id11/exec/processes/UMC> .34.1  <

trying to get source for 10.65.34.1

source should be 10.80.4.15

traceroute to 10.65.34.1 (10.65.34.1) from 10.80.4.15 (10.80.4.15), 30 hops max

outgoing MTU = 1500

1  10.80.4.3 (10.80.4.3)  1 ms  0 ms  0 ms

2  172.22.2.9 (172.22.2.9)  1 ms  0 ms  0 ms

3  10.32.1.6 (10.32.1.6)  0 ms  0 ms  0 ms

4  10.71.242.62 (10.71.242.62)  3 ms  3 ms  3 ms

5  * * *

6  * * *

7  * * *

8  * * *

9  * * *

10  * * *

11  * * *

12  * * *

13  * * *

14  * * *

15  * * *

16  * *

[Russ Ross]

It is fortunate you showed the MTU=1500 setting because that can be hard to uncover and certainly can be the source of routers hanging.

I recently had to make this very painful discovery when my DSL modem and Linksys router were both set to MTU=1500 by default, causing them to hang when visiting web sites (ex: http://www.mozilla.com) because some responses back are adding an extra 8 bytes for added security these days, which hangs communications with routers with MTU=1500.

At least this is the explanation I got and setting MTU=1492 on all the pathways solved all my problems with hanging.

I’m not saying this will solve your problem but I strongly reccommend you set MTU=1492 because it will cause things to hang sooner or later and is a hidden time bomb waiting to go off.

I would be interested to know if hanging still occurs after you set MTU=1492, assuming that can be accomplished by you or a network guy.

Russ Ross
RussRoss318@gmail.com

[Yvonne Gaffney]

The vendor we are connecting with has a tcp max segment size of 960.  Is  there a way for us to modify the max segment size on the Cloverleaf?

[John Mercogliano]

Yvonne, you don’t want to do that.

John Mercogliano
Sentara Healthcare
Hampton Roads, VA

[Author]

Replies