This is the mail archive of the mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

ramped CPU problem with fetchmail

Output from:
fetchmail --version:
This is fetchmail release 6.1.2+NTLM+SSL+NLS
Fallback MDA: /usr/bin/procmail -d %T
CYGWIN_NT-5.1 CWEEKS-LAP 1.3.15(0.63/3/2) 2002-11-07 13:57 i686 unknown
Taking options from command line and /home/cpw/.fetchmailrc
Logfile is /var/log/fetchmail.log
Idfile is /home/cpw/.fetchids

Other info is attached.

Problem summary:

It appears that the CPU ramp problem with emacs that has been discussed on this mailing list is not restricted to emacs. Fetchmail, running in daemon mode, reliably exhibits the same behaviour. The exact setup is that the fetchmail daemon is NOT able to successfully look up the name of my mailhost (it's inside a VPN that isn't dialled up when the spin happens). The fetchmail daemon, whether run as fetchmail -d or as a service will then start consuming 99-100% CPU until it is killed. It should be noted that if it can successfully get the DNS name of the mail server it doesn't exhibit this behaviour (until it can't hit it).

Attached are strace and cygcheck output. It appears that the WM_TIMER signal sets the process spinning out of control. Maybe the winsock is in some kind of invalid state because of the dns failure and makes the timer fail? Pure speculation.


Attachment: cygcheck.out
Description: Text document

Attachment: trace.log
Description: Text document

Unsubscribe info:
Bug reporting:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]