T
TDC
Anyone know why this would fail?
Note that the log file timestamps shows that these all happened witin
the same second!
=======================================================
Dim lMillisElapsed As Long
Dim lStartTicks As Long = DateTime.Now.Ticks
mTracer.Trace("ReceiveMsg::RecvTimeout is " & iRecvTimeout)
LOG READS> ReceiveMsg::RecvTimeout is 60000
While lMillisElapsed < iRecvTimeout
mTracer.Trace("ReceiveMsg::MillisElapsed = " & lMillisElapsed)
LOG READS> ReceiveMsg::MillisElapsed = 0
'DATA READ OFF A SOCKET
Dim lCurrTicks As Long = DateTime.Now.Ticks
mTracer.Trace("ReceiveMsg::CurrTicks is " & lCurrTicks)
LOG READS>ReceiveMsg::CurrTicks is 632828638200000000
lMillisElapsed = (lCurrTicks - lStartTicks) \
TICKS_PER_MILLISECOND
mTracer.Trace("ReceiveMsg::MillisElapsed (2) = " &
lMillisElapsed)
LOG READS> ReceiveMsg::MillisElapsed (2) = 3600000
'AT THIS POINT THE 3600000 (which is 1 hour) CAUSES An ERRONEOUS
TIMEOUT
=======================================================
TICKS_PER_MILLISECOND is declared as in Integer (Int32), but I can't
see how that would be a problem. This problem happens very
intermittenly and it is always an hour (3600000) when it happens. SQL
CE is being populated by the data being communicated and as it happens,
if we comment out all DB code we don't seem to hit this (or perhaps not
as often anyway). In fact, just adding the trace code (which appends
to a ArrayList of strings, not to a file) seemed to have reduced these
false timeouts.
Any advice is appreciated.
Tom
Note that the log file timestamps shows that these all happened witin
the same second!
=======================================================
Dim lMillisElapsed As Long
Dim lStartTicks As Long = DateTime.Now.Ticks
mTracer.Trace("ReceiveMsg::RecvTimeout is " & iRecvTimeout)
LOG READS> ReceiveMsg::RecvTimeout is 60000
While lMillisElapsed < iRecvTimeout
mTracer.Trace("ReceiveMsg::MillisElapsed = " & lMillisElapsed)
LOG READS> ReceiveMsg::MillisElapsed = 0
'DATA READ OFF A SOCKET
Dim lCurrTicks As Long = DateTime.Now.Ticks
mTracer.Trace("ReceiveMsg::CurrTicks is " & lCurrTicks)
LOG READS>ReceiveMsg::CurrTicks is 632828638200000000
lMillisElapsed = (lCurrTicks - lStartTicks) \
TICKS_PER_MILLISECOND
mTracer.Trace("ReceiveMsg::MillisElapsed (2) = " &
lMillisElapsed)
LOG READS> ReceiveMsg::MillisElapsed (2) = 3600000
'AT THIS POINT THE 3600000 (which is 1 hour) CAUSES An ERRONEOUS
TIMEOUT
=======================================================
TICKS_PER_MILLISECOND is declared as in Integer (Int32), but I can't
see how that would be a problem. This problem happens very
intermittenly and it is always an hour (3600000) when it happens. SQL
CE is being populated by the data being communicated and as it happens,
if we comment out all DB code we don't seem to hit this (or perhaps not
as often anyway). In fact, just adding the trace code (which appends
to a ArrayList of strings, not to a file) seemed to have reduced these
false timeouts.
Any advice is appreciated.
Tom