Re: reconnect failing in 3.51.12

Stay-alive messages definitely have their place.

I have a serial port device that times out after one _second_ of
inactivity, so I use a timer to send it a specific message to keep the
port open.

Re: the process' crashing, perhaps very few people try to keep a
connection open for so long, so fixing the bug isn't worth the effort .
. . especially since a workaround is to "do nothing" once in a while.

-- Jordan Bradford
----------------------------
BrownBoots Interactive, Inc.
108 South Main St.
Fond du Lac, WI 54935
920-906-9175
http://www.brownboots.com/


John Calcote wrote:
> Oh, I completely understand the necessity for the timeout. Frankly, I'm
> surprised the default isn't more like 30 minutes rather than 8 hours. My
> problem is that I can't seem to (manually) reconnect within the process
> after a timeout without corrupting the process - ultimately causing
> either a failure to reconnect, or a segfault (and subsequent crash).
>  
> John
>  
> -----
> John Calcote ([email protected])
> Sr. Software Engineeer
> Novell, Inc.
> 
> 
>>>> On Sat, Feb 10, 2007 at  7:04 PM, in message
> <003701c74d80$ef84d100$6500a8c0@winxppro>, "Lawson Cronlund"
> <[email protected]> wrote:
> 
> John,
> 
> My interpretation was that the no-activity timeout operation is actually
> a protection that should be in the MySQL server.  It's a protection that
> retires connections that are not properly terminated.  If nothing else
> was done, in a poorly implemented system, connections might be left
> "hanging" with random failures on maximum allowed connections.
> 
> The documentation seems to say that this timeout interval can be set in
> the configuration file(s) but any changes that I made didn't seem to
> affect the timeout interval - so I gave up and did the dirty
> implementation which could be viewed as a sort of watchdog to tell me if
> my application was messing up if the connection ever terminated
> unilaterally.
> 
> The result was that my application was improved since it led me to using
> the technique for the watchdog program in my application that monitors
> the alive/dead status of the other application programs in the system.
> 
> I guess what I'm saying is that I think that the no-activity connection
> timeout in MySQL is a good thing but I would have liked it if I could
> have found the method for changing the timeout value.
> 
> Regards,
> 
>  
> 
> 
> ------------------------------------------------------------------------
> 
> BEGIN:VCARD
> VERSION:2.1
> X-GWTYPE:USER
> FN:John Calcote
> TEL;WORK:1-801-861-7517
> ORG:;Unified Identity System Eng TE
> TEL;PREF;FAX:801/861-2292
> EMAIL;WORK;PREF;NGW:[email protected]
> N:Calcote;John;;Sr. Software Engineer
> TITLE:Sr. Software Engineer
> ADR;DOM;WORK;PARCEL;POSTAL:;PRV-H-511;;Provo
> LABEL;DOM;WORK;PARCEL;POSTAL;ENCODING=QUOTED-PRINTABLE:John Calcote=0A=
> PRV-H-511=0A=
> Provo
> END:VCARD
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> 

-- 
MySQL ODBC Mailing List
For list archives: http://lists.mysql.com/myodbc
To unsubscribe:    http://lists.mysql.com/[email protected]