L
Leopold
Very honoured Mr. Felbermayr!
Due to unserers discussion on past Sunday us some ideas
are concerning the problem "Patriot hang themselves up
sporadically" come. After somewhat dealing analysis of the
problem the reason of the problem lies quite actually
clearly with the hand. A cause of these again and again,
sporadically and without to connection of arising crashes
are to be found in the data base structure, or more
exactly said, in the underlying data base engine - ACCESS
2,0 -. All data are stored into ACCESS files at the file
server. The structure of the railways corresponds to the
possibilities of ACCESS 2,0 ACCESS 2,0 has now some at
inability which data base by run gene concerns. If someone
calls station up data at a Patriot, then in the data base
ranges of the tables will change closed thereby no second
accesses the data for the moment the access can. One calls
this data base blockage Pagelocking. Due to the
restrictions of ACCESS 2,0 relatively large ranges become
closed with the Pagelocking. Footstep now the situation on
that someone to the same time as the first user a Patriot
station used AND the data those the second user call
up/worked on in the same PacgeLocking range is appropriate
as those of the first user occurs for "DEADLOCK" a
situation! DEADLOCK means that at least two processes
access (almost) the same data base range. Due to the first
Pagelockings no second Pagelocking of the same range can
be made more and for safety reasons (which not go
aroundable and switch offable are) occurs this total
blockage of the data base! This procedure/mechanism shows
up as to then hang remain the Patriot stations. Why was
there this problem not already in former times? This
problem always gave it, however the probability is that a
deadlock arises on the Nutzungsintensivitaet on Patriot
exclusively dependent. By the rise of the number of
employees as well as by the increase of the number of
Patriot stations naturally also the number of the accesses
and concomitantly the probability deadlock of a situation
rise. It is causal always the inability that is the basis
data base engine ACCESS 2,0 those a relative large data
base range always closes. Only starting from the version
ACCESS 97 was halfway defused on the part of Microsoft the
problem. No being to blame for is thus present on the part
of Patriots! Solutions: A solution was already mentioned:
armament on ACCESS 97. However is to assume that that due
to the large version jump of ACCESS 2,0 on ACCESS 97
Patriot will not any longer function. An alternative
employment is to convert the ACCESS 2,0 data base after
SQL. Itself e.g. MySQL or MS-SQL would offer. Not only the
improved speed behavior of a SQL of server would become
positively apparent; also the substantially more
intelligent Lockingmechanismen than with ACCESS 2,0 and
also ACCESS 97/2000/XP would create the problem with high
probability from the world! Conclusio: The cause of the
problems are to be searched in the inability of ACCESS 2,0
and/or the risen use intensity from Patriot to. Since we
will hardly decrease the use intensity to be able are the
beginning, which is converted inability by ACCESS 2,0 to
eliminate in that the data base after SQL! We are the
opinion, which is this scenario described here to 97% the
cause for "crashes" from Patriot. Due to that vorhandnen
practical experiences (Patriot in this version is not our
knowledge otherwise anywhere in the use) we can naturally
deliver no warranty in jedwender kind! But we see the
chances of success quite positive.
Due to unserers discussion on past Sunday us some ideas
are concerning the problem "Patriot hang themselves up
sporadically" come. After somewhat dealing analysis of the
problem the reason of the problem lies quite actually
clearly with the hand. A cause of these again and again,
sporadically and without to connection of arising crashes
are to be found in the data base structure, or more
exactly said, in the underlying data base engine - ACCESS
2,0 -. All data are stored into ACCESS files at the file
server. The structure of the railways corresponds to the
possibilities of ACCESS 2,0 ACCESS 2,0 has now some at
inability which data base by run gene concerns. If someone
calls station up data at a Patriot, then in the data base
ranges of the tables will change closed thereby no second
accesses the data for the moment the access can. One calls
this data base blockage Pagelocking. Due to the
restrictions of ACCESS 2,0 relatively large ranges become
closed with the Pagelocking. Footstep now the situation on
that someone to the same time as the first user a Patriot
station used AND the data those the second user call
up/worked on in the same PacgeLocking range is appropriate
as those of the first user occurs for "DEADLOCK" a
situation! DEADLOCK means that at least two processes
access (almost) the same data base range. Due to the first
Pagelockings no second Pagelocking of the same range can
be made more and for safety reasons (which not go
aroundable and switch offable are) occurs this total
blockage of the data base! This procedure/mechanism shows
up as to then hang remain the Patriot stations. Why was
there this problem not already in former times? This
problem always gave it, however the probability is that a
deadlock arises on the Nutzungsintensivitaet on Patriot
exclusively dependent. By the rise of the number of
employees as well as by the increase of the number of
Patriot stations naturally also the number of the accesses
and concomitantly the probability deadlock of a situation
rise. It is causal always the inability that is the basis
data base engine ACCESS 2,0 those a relative large data
base range always closes. Only starting from the version
ACCESS 97 was halfway defused on the part of Microsoft the
problem. No being to blame for is thus present on the part
of Patriots! Solutions: A solution was already mentioned:
armament on ACCESS 97. However is to assume that that due
to the large version jump of ACCESS 2,0 on ACCESS 97
Patriot will not any longer function. An alternative
employment is to convert the ACCESS 2,0 data base after
SQL. Itself e.g. MySQL or MS-SQL would offer. Not only the
improved speed behavior of a SQL of server would become
positively apparent; also the substantially more
intelligent Lockingmechanismen than with ACCESS 2,0 and
also ACCESS 97/2000/XP would create the problem with high
probability from the world! Conclusio: The cause of the
problems are to be searched in the inability of ACCESS 2,0
and/or the risen use intensity from Patriot to. Since we
will hardly decrease the use intensity to be able are the
beginning, which is converted inability by ACCESS 2,0 to
eliminate in that the data base after SQL! We are the
opinion, which is this scenario described here to 97% the
cause for "crashes" from Patriot. Due to that vorhandnen
practical experiences (Patriot in this version is not our
knowledge otherwise anywhere in the use) we can naturally
deliver no warranty in jedwender kind! But we see the
chances of success quite positive.