Load Error in Dump and Load by Data Adminstration
- Thread starter wolfskin
- Start date
RealHeavyDude
Well-Known Member
AFAIK the ASCII dump & load had that limitation back then. Any Progress V9 is stone age software, I can only recommend you to upgrade to a reasonably recent version of OpenEdge.
In the meantime you can work around that problem doing a binary dump and load which also usually is much faster.
Heavy Regards, RealHeavyDude.
In the meantime you can work around that problem doing a binary dump and load which also usually is much faster.
Heavy Regards, RealHeavyDude.
Do binary dump and load is same function of ASCII dump and load, I read from some source that after ASCII dump and load, all data will become in sequence and index will rebuild again, but binary dump and load will not change anything, so the performance of database will not change after done it. Is it really?
RealHeavyDude
Well-Known Member
No - that is not correct.
The binary dump dumps the data as is but it does it in sequence - usually it picks the primary index of the table, but you can override that behavior. That is basically the same behavior than the ASCII dump expect that it dumps the data using the code page defined via -cpstream in ASCII format whereas the binary dump does not do any code page conversion. The binary load deactivates all indexes on the table and then loads the table with no index activated. That means that you must run an index rebuilt for the loaded table after the binary load has finished or you can direct the binary load to do that automatically - which one the best is is subject to discussion.
Heavy Regards, RealHeavyDude.
The binary dump dumps the data as is but it does it in sequence - usually it picks the primary index of the table, but you can override that behavior. That is basically the same behavior than the ASCII dump expect that it dumps the data using the code page defined via -cpstream in ASCII format whereas the binary dump does not do any code page conversion. The binary load deactivates all indexes on the table and then loads the table with no index activated. That means that you must run an index rebuilt for the loaded table after the binary load has finished or you can direct the binary load to do that automatically - which one the best is is subject to discussion.
Heavy Regards, RealHeavyDude.
TomBascom
Curmudgeon
There are several persistent myths about binary dump that ought to be debunked:
Myth: Binary dump is not portable across different OS, byte-orders or versions of Progress.
Fact: Binary dump *is* portable between operating systems, across byte orders and to higher versions of Progress (you cannot, however, expect to go backwards).
Myth: Binary dump is "block level".
Fact: Binary dump is a record level operation.
Myth: Binary dump & load does not reorganize data.
Fact: Binary dump & load reorganizes data just as effectively as any other dump & load.
Myth: Binary dump is not portable across different OS, byte-orders or versions of Progress.
Fact: Binary dump *is* portable between operating systems, across byte orders and to higher versions of Progress (you cannot, however, expect to go backwards).
Myth: Binary dump is "block level".
Fact: Binary dump is a record level operation.
Myth: Binary dump & load does not reorganize data.
Fact: Binary dump & load reorganizes data just as effectively as any other dump & load.