[WriteLog] writelog sticks with a lot of qsos

[Kostas SV1DPI]

Tnx all for the answers. I don't have other programs in the computers as 
I use them just for contesting. One of them is running the windows 
antivirus program. I remember that when the problem happened the last 
year, I tried and deactivated the antivirus without any results.
What Mike N7MH says it makes sense to me. This describes the problem and 
fits in my situation. I remember that, in several cases, I have found 
one of the computers with the network status window blinking. I will 
notice this detail better in the future.
I don't think is rf in the network also because I use a professional 
cisco switch and router. My ethernet cables also are professional made. 
Plus on these, the network works fine without any brake when I run wintest.
Unfortunately I can not try to reproduce it immediately as the contest 
shack is not in the home but I will try it soon and let you to learn my 
results.

73 Kostas SV1DPI

On 30/9/2015 9:00 μμ, Mike Heideman wrote:
> Since Kostas is using 3 networked WriteLog instances his symptoms may 
> be related to a network disconnect issue that I described 3 years 
> ago.  If his home computer is not networked to another WriteLog 
> instance then that may explain why it isn't reproducible there.
>
> http://lists.contesting.com/_writelog/2012-11/msg00113.html
>
> If any of the computers, even the one being used exclusively for 
> backup and clusters, is temporarily disconnected from the network then 
> keying delays start to occur on one or more of the other 
> computers and get progressively worse, even after reconnecting the 
> network.  Exiting and restarting WriteLog on the affected computer 
> eliminates the delays.  This happened to me in a Sprint contest with 
> only a small number of QSOs logged so it doesn't happen just when a 
> thousand or more QSOs have been logged.
>
> Keyboard response is delayed for all typing.  Callsigns that are typed 
> into the input Call field are echoed only after a long delay.  
> Function keys or mapped keys that send messages are similarly delayed.
>
> I don't have a logging computer at home so I can't verify this now, 
> but here's how I believe that this problem can be reproduced.
>
> Start up WL for some contest on two computers.  Register to accept 
> network connections on both computers.  Link to network from one 
> computer to the other.  Log some QSOs from each computer (this may not 
> be necessary).  Disconnect one of the computers from the network 
> either by removing the network cable, disconnecting Wi-Fi, or similar 
> hard disconnect.  Log some QSOs on the other computer, using 
> function/mapped keys to send messages.  After 15 minutes, possibly 
> less, there will be noticeable delays in echoing of characters and in 
> sending messages.  The network can be reconnected and the WL instances 
> relinked but the keyboard delays still persist on the one computer.
>
> Hopefully this issue can be fixed in WL as it has happened to me many 
> times, several of which I've documented in my earlier posting.  It was 
> only when it happened during Sprint with no cluster and a nearly empty 
> log that I realized the broken network connections were the common 
> thread between all the occurrences.
>
> 73,
> -Mike, N7MH
>
> Dean/8P6SH wrote:
> > Kostas
> >
> > Not an expert by any means but I have had similar problems. Usually 
> it is
> > related to the combination of memory, processor and other apps that 
> might
> > be running on the same computer as well as possibly network load 
> though I
> > cannot be 100% sure.
> >
> > In my situation there was an issue with drivers and the operating system
> > hogging resources and not returning unused memory. This I believe is why
> > you see things get better after a reboot and get bad again eventually.
> >
> > I would try not to run stuff I don't need and keep configurations 
> simple.
> > Also if you have unused ports in your microham keyer device, you 
> should try
> > deleting them - the virtual ports that are unused.
> >
> > If you go through your systems systematically I'm sure you will find 
> things
> > running that could/should be turned off when contesting.
> >
> > In my case another culprit was anti-virus. I turn mine off during 
> contests
> > but I also don't visit questionable sites while the system is not 
> connected.
> >
> > Hope you solve the problem. Writelog is great but of course windows does
> > have its issues.
> >
> > CHeers
> >
> > Dean - 8P6SH
> >
> >
> >
> >
> >
> >
> > *Dean St.Hill*
> >
> >
> >
> >
> > *Barbados.JustAskLocals.com <http://Barbados.JustAskLocals.com>*
> > *It's A Big World. Travel Smarter <http://barbados.justasklocals.com/>*
> >
> > *17 Ocean City*
> > *St.Philip*
> > *Barbados*
> >
> > *246-825-2039*
> > *246-416-5277*
> >
> > *dmsthill at gmail.com <dmsthill at gmail.com>*
> >
> > On Tue, Sep 29, 2015 at 5:27 PM, Kostas SV1DPI <sv1dpi at otenet.gr> wrote:
> >
> > > I have noticed that writelog sticks momentarily sometimes. I have 
> noticed
> > > it happening after 1000 qsos in the log. It happens on cw and 
> rtty. On cw
> > > while I hit the function key, I am waiting for 1-2 seconds to start
> > > sending. On rtty it starts keying something like RY for a while 
> and after
> > > 1-2 seconds the message starts. It happens after some hours. A 
> fresh reboot
> > > corrects the problem for a while but it happens again. The result 
> of this
> > > is that I use the wintest the last time for cw contests but I prefer
> > > writelog. The biggest problem is that I noticed it on rtty also 
> the last
> > > time...
> > > Is anyone other who has noticed this behaviour?
> > > The keying device is microham in the first radio and winkey plus 
> homemade
> > > interface to the 2nd radio (I don't think the problem is in 
> hardware as it
> > > works fine with wintest and dx4win). Computers are in network and 
> I use the
> > > 11.28 version but I have noticed it in all of the last versions. 
> Computers
> > > in the shack are duo-core with 2gb memory and windows 7. A 3rd 
> laptop is
> > > used for backup and runs the clusters.
> > > I can not reproduce it in my home computer. I believe this is why the
> > > computer is not so many hours on or the clusters spots are no so 
> many. My
> > > first thought was that heavy spots from the clusters was the cause 
> but on
> > > rtty I had not skimmer on and the spots were much lower. I talked 
> it with a
> > > friend of mine who runs a much older version of writelog and a new 
> computer
> > > with lot of memory and he has noticed it also. This is why I am 
> asking for
> > > comments...
> > >
> > > 73 Kostas SV1DPI
>