I was running a single pair of poly machines, and after a few hours, during
the fill phase, my polyclt dies with the message:
Virtual memory exceeded in 'new'
I grepped through the polygraph source and didn't see this, so I'm assuming
that it may be a FreeBSD message, and perhaps occurs when creating a new
object. Any ideas about why this would occur? Is there perhaps a memory
leak that is causing this? I'm not seeing any other errors, except for a
couple "premature end of msg body" errors. The first one occurs during
i-framp and the other during i-fill. But the polyclt dies at the 161.12
mark, which is (curiously) still in the i-fill phase.
...
007.34| i-framp 18794 53.37 2665 4.23 0 3429
007.40| ./Xaction.cc:79: error: 1/1 (263) premature end of msg body
965225227.398392# obj:
http://10.200.129.12/w0326df02.342e2355:0000063a/t04/_000
00001 xact: 0326df02.342e2355:00009d0e
[no data to dump]
007.43| i-framp 19092 59.78 2691 5.41 1 3429
...
077.47| i-fill 465181 107.99 36677 3.96 0 4000
077.49| ./Xaction.cc:79: error: 2/2 (263) premature end of msg body
965229432.335441# obj:
http://10.200.130.106/w0326df02.342e2355:00000566/t04/_00
000001 xact: 0326df02.342e2355:000e3c34
[no data to dump]
077.55| i-fill 465722 108.29 36756 3.22 1 4000
...
Brian O'Keefe (brianok@us.ibm.com)
IBM Transarc Lab, Pittsburgh, PA
This archive was generated by hypermail 2b29 : Tue Jul 10 2001 - 12:00:14 MDT