I have a big problem

General Help regarding HMG, Compilation, Linking, Samples

Moderator: Rathinagiri

martingz
Posts: 395
Joined: Wed Nov 18, 2009 11:14 pm
Location: Mexico

Re: I have a big problem

Post by martingz »

We continue to have problems, but we already detected what is happening, for example, Box 1 collects 200 receipts and makes a cut, until then everything is fine, it continues to collect with the same date, but when taking the second cut of the same day, only the receipts that I collected appear. After the first cut and the latter are not reflected in the server's databases, we restart the application and it continues to show only the last cut (7 receipts), this never reaches the server, but they do exist in Box 1 (there are no databases). data on the local computer), just in case it helps someone.
hansmarc
Posts: 40
Joined: Thu Jun 23, 2016 5:38 am
Location: Belgium

Re: I have a big problem

Post by hansmarc »

Hi Martingz,

simply an idee, what size have the BIG tables.
There is a problem with "record locking" if your tables are bigger as 1GB.
You have to set another locking scheme if this is the case.
Might be the cause of your problems .....

We had the same kind of problem 'loss of data' some years ago, took me months to find the cause.
Client had DBFCDX/1010 errors with DOS ERROR 33 message with loss of data as result.

Below a small table and some info to change the lockingscheme.
We set it to "4" and didn't see any problems for years.

Regards,
Hans

Locking schemes Val. Description
--------------- ---- ------------------------------------------------------------------------
DEFAULT 0 CLIPPER default locking scheme
COMIX 2 COMIX and CL53 DBFCDX hyper locking scheme
VFP 3 [V]FP, CL52 DBFCDX, SIx3 SIXCDX, CDXLOCK.OBJ
HB32 4 table size max 4GB, Harbour hyper locking scheme for 32-bit file API
HB64 5 table size no limit, Harbour hyper locking scheme for 64-bit file API,
CLIPPER2 6 extended Clipper locking scheme NTXLOCK2.OBJ

Function SET(_SET_DBFLOCKSCHEME, 0...6 )
martingz
Posts: 395
Joined: Wed Nov 18, 2009 11:14 pm
Location: Mexico

Re: I have a big problem

Post by martingz »

Hansmarc

The dia_pag table is very small, it collects all payment transactions, at the end of the day, an application is executed that distributes the data between 2 different departments, billing and accounting, then everything is deleted and a pack is given and it is like new.
The problem is that on the computers it looks as if another table existed on the local computer, which does not exist.
User avatar
AUGE_OHR
Posts: 2065
Joined: Sun Aug 25, 2019 3:12 pm
DBs Used: DBF, PostgreSQL, MySQL, SQLite
Location: Hamburg, Germany

Re: I have a big problem

Post by AUGE_OHR »

hi,

have you check SMB Setting :!:

your old SMB Setting, for Windows XP, will not work well under Windows 10 / 11
have fun
Jimmy
martingz
Posts: 395
Joined: Wed Nov 18, 2009 11:14 pm
Location: Mexico

Re: I have a big problem

Post by martingz »

Auge
All our client computers are Windows 10, the Windows XP thing was a test that we used to see how the same systems behaved on that computer, the strangest thing is on my Windows 11 computer I make the information pass to the larger tables and everything is Good, but on a computer with Windows 10 it does not pass all the records, even if it passes it several times, without doing a reindex or anything like that, on my computer on the same tables it runs perfectly
Pcmodula
Posts: 34
Joined: Thu Nov 22, 2012 6:00 pm

Re: I have a big problem

Post by Pcmodula »

Ola Martin,
one of the problems encountered with Windows 10 (and I think also Windows 11) is related to the fact that a write to disk is not immediately performed, so a possible dbcommit() returns .T. but in fact he hasn't physically written yet.
So think about the chaos that is generated if you read, from the hard disk ,a table that you think is up to date.
I have solved this trouble , index corruption, slow network performance using Netio and then I switched to Mysql.
Is only a my experience not a text bible.
My two cents
Saludos
Pcmodula
martingz
Posts: 395
Joined: Wed Nov 18, 2009 11:14 pm
Location: Mexico

Re: I have a big problem

Post by martingz »

Pcmodula
We are changing all our systems to MariaDB, but there are too many, we have already finished Payroll, Warehouse and purchasing, Suppliers, Vehicles, but we are missing the largest and most complex ones, which are the Commercial and the accounting, these last 2 go hand in hand, They have to leave at the same time

Thank you very much for your comments and I think that is where the issue lies, after the dbcommit() I put a do events and there were up to 3 duplicate records.
User avatar
serge_girard
Posts: 3170
Joined: Sun Nov 25, 2012 2:44 pm
DBs Used: 1 MySQL - MariaDB
2 DBF
Location: Belgium
Contact:

Re: I have a big problem

Post by serge_girard »

Pity you could not find the cause...!
There's nothing you can do that can't be done...
martingz
Posts: 395
Joined: Wed Nov 18, 2009 11:14 pm
Location: Mexico

Re: I have a big problem

Post by martingz »

Serge
No
We are dedicating ourselves more to finishing the migration to MariaDB, we made some applications to facilitate the passage of information, and here is the strangest thing, an Accounting computer (Windows 10) that is used for the passage of information, does not process all the payments, you can do it several times and the result is the same, on my computer (Windows 11) it carries out the process without any problem
User avatar
serge_girard
Posts: 3170
Joined: Sun Nov 25, 2012 2:44 pm
DBs Used: 1 MySQL - MariaDB
2 DBF
Location: Belgium
Contact:

Re: I have a big problem

Post by serge_girard »

Windows misery ...
There's nothing you can do that can't be done...
Post Reply