Bugs item #1357294, was opened at 2005-11-15 12:37
Message generated for change (Comment added) made by mlkersten
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=1357294&group_id=56967

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Mapi
Group: (zombie: MonetDB 4.8.0)
>Status: Closed
>Resolution: Out of Date
Priority: 5
Private: No
Submitted By: Fabian (mr-meltdown)
Assigned to: Fabian (mr-meltdown)
Summary: MAPI: protocol violation on GDKmalloc failure

Initial Comment:
When the client is in a session with the server, and
the server hits a GDKmalloc failure, this message is
sent to the server as is (i.e. without error prefix (!).

Example message:
GDKmalloc(4194304) fail => BBPtrim(enter) usage
[mem=634350178,vm=67108864]

----------------------------------------------------------------------

>Comment By: Martin Kersten (mlkersten)
Date: 2007-01-20 23:48

Message:
Logged In: YES 
user_id=490798
Originator: NO

NO TEST NEEDED
There have been no reported errors and further code developments
should continuously be aware of proper marking errors.

----------------------------------------------------------------------

Comment By: Fabian (mr-meltdown)
Date: 2006-11-05 22:00

Message:
Logged In: YES 
user_id=963970

That's fine with me, and should work nicely with the latest
changes to the protocol.

----------------------------------------------------------------------

Comment By: Martin Kersten (mlkersten)
Date: 2006-11-05 20:37

Message:
Logged In: YES 
user_id=490798

It seems the 200 cases in GDK can be handled, provided we have
a clear policy. Now, debugging/error messages are sent both to
GDKout and GDKerr.
I propose to sent output intended for GDKout with a '#'
and all for GDKerr with a '!'
This way trace information also goes to MapiClient, rather
then the server console.
It is up to the MapiClient implementation(or front-end) to
filter the relevant data.
  

----------------------------------------------------------------------

Comment By: Martin Kersten (mlkersten)
Date: 2006-06-11 18:40

Message:
Logged In: YES 
user_id=490798

Checked gdk_utils.mx for direct THRprintf that should have
included a ! to designate an error.
Check results in testweb tomorrow.

----------------------------------------------------------------------

Comment By: Martin Kersten (mlkersten)
Date: 2006-06-06 23:35

Message:
Logged In: YES 
user_id=490798

The specific errors are easily trapped in gdk_utils.mx.
The real problem is that THRprintf is called in many places
 subject to a debugging flag being set. This may cause
'garbage' to be sent over the line.
A total of around 300 THRprintfs should be checked.

It takes a few hours to inject '!' in all places.

----------------------------------------------------------------------

Comment By: Fabian (mr-meltdown)
Date: 2006-06-06 21:09

Message:
Logged In: YES 
user_id=963970

should be on our todo's again...

----------------------------------------------------------------------

Comment By: Niels Nes (nielsnes)
Date: 2006-01-21 00:06

Message:
Logged In: YES 
user_id=43556

Currently there is no warning/debugging header, such as the
errors '!'. We first need to introduce such a character and
change/check all THRprint calls. Lowering the priority.
Assigning to fabian for defining this change to mapi.

----------------------------------------------------------------------

Comment By: Niels Nes (nielsnes)
Date: 2005-12-20 08:08

Message:
Logged In: YES 
user_id=43556

BugDay_2005-12-19, Niels: NO TEST / 
Protocol is checked by any mapi/jdbc client

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=1357294&group_id=56967

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Monetdb-bugs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-bugs

Reply via email to