cancel
Showing results for 
Search instead for 
Did you mean: 

ace overflow

Michael Schulte zur Sur
Honored Contributor

ace overflow

Hi all,

we have a ds25 with 5.1b pk4 with a fax machine attached to the serial port. Whenever we have high I/O on our sa5300 storage unit, we get the error
Mar 21 14:57:26 dus10d vmunix: ace1: input silo overflow; status = 63
and the fax program terminates.
Are there other system parameter besides in the subsystem ACE that may influence the performance?
What does the debug parameter do?

Thanks for infos,

Michael
9 REPLIES
Ann Majeske
Honored Contributor

Re: ace overflow

Hi Michael,

I suggest that you open a case with Tru64 UNIX Support on this.

Ann
Michael Schulte zur Sur
Honored Contributor

Re: ace overflow

Hi Ann,

thanks for posting. That is what we did. HP replaces the motherboard but the problem persists.

greetings,

Michael
Johan Brusche
Honored Contributor

Re: ace overflow


Michael,

First check with consvar that com1_flow and com2_flow are set to SOFTWARE, ie consvar -g com2_flow. If the cabling and faxdevice at the other end provide RTS/CTS signal capability you can set com2_flow to HARDWARE or BOTH

sysconfig -q ace , shows you that the ace major number is 35.
Now do:

cd /dev
ls -al | grep " 35,"
You will find tty00 and tty01, tyy01 is probably the one , the fax program uses.

stty -a -f /dev/tty01, will give you the settings.
If the fax-device at the other end is capable of reacting to XOFF, then the tty01 should have ixoff (not -ixoff)

You probably will not be able to change the setting(s) with stty as long as the fax program is owner of the device. Normally the fax-program has the responsability to set parameters via ioctl and tcsetattr.

Regards,
Johan.

_JB_
Michael Schulte zur Sur
Honored Contributor

Re: ace overflow

Johan,

thanks for posting. The parameter is -ixoff and the com srm parameter are
com1_baud =
com1_modem =
com1_flow = 9600
com2_baud = OFF
com2_modem = SOFTWARE
com2_flow = 9600

I have set com2_mode to hardware and ixoff.

Are you sure, you did not mistake the flow?

greetings,

Michael
Michael Schulte zur Sur
Honored Contributor

Re: ace overflow

Johan,

the error persists.

greetings,

Michael
Johan Brusche
Honored Contributor

Re: ace overflow

Michael, there is a mix-up somewhere

You show

com2_baud = OFF
com2_modem = SOFTWARE
com2_flow = 9600

It should be

com2_baud = 9600
com2_modem = OFF (ON if modem signals)
com2_flow = SOFTWARE
Above items are console SRM >>> params
that can be examined(only) via consvar, but to be set at HALT prompt P00>>>

The ixoff parameter is an OS setting for the tty driver.


Rgds,
__ Johan.

_JB_
Michael Schulte zur Sur
Honored Contributor

Re: ace overflow

Johan,

this machine has got a new motherboard because of this problem and the firmware is now 7.1-2. The technician did not know about the heap expansion varaiable.
Are there more known issues with that?

thanks,

Michael
Michael Schulte zur Sur
Honored Contributor

Re: ace overflow

Johan,

I tried to get some information in the internet concerning these console variables but even the doc for the ds25 srm does not tell too much about it. For examples, what is com2_misc?
com1_baud = 9600
com1_modem = OFF
com1_flow = HARDWARE
com2_baud = 9600
com2_modem = OFF
com2_flow = HARDWARE

Is this configuration setting ok?

thanks,

Michael
Michael Schulte zur Sur
Honored Contributor

Re: ace overflow

Johan,

this is not the only machine with that type of srm variable hickup. I have got another DS20 and a DS25 with that problem.

Michael