Copy 100GB MSSQL 9.0 Database to Oracle 10g
in [Oracle]
|
Prev: ORA- in alert
Next: ORA-00604/ORA-16000 on standby
From: Mladen Gogala on 13 Jun 2010 17:14 On Sun, 13 Jun 2010 20:51:44 +0000, Terry Dykstra wrote: > You can use InfoMaker (Sybase). It has a so called pipeline utility > which allws you to move data between ant 2 databases. Very easy to use. > > Terry Dykstra My favorite utility for complex cases is still Perl although I am reading a Python book and I like what I read. If only there was a comprehensive Python archive network, with tons of modules and applications available, Python would immediately take over the world by storm. PyPI is not nearly as good - yet. -- http://mgogala.byethost5.com
From: if_investor on 14 Jun 2010 00:20 Hi, Thanks for the replies. Some people asked me for more details about the issue. There is an MSSQL 9.00 database about 100GB in size that we need to get a copy of; To be more accurate, we need to get hold of its major schema, and create an Oracle database (Hoping that it will be 10g) with the same schem, having the same data. (The entire database consists of only that schema and whatever the default MSSQL schemas are). I am not sure if the source DB has unique MSSQL features; I don't know MSSQL, and I am not sure sure what to check. Thanks, QZ On Jun 13, 5:14 pm, Mladen Gogala <gogala.mla...(a)gmail.com> wrote: > On Sun, 13 Jun 2010 20:51:44 +0000, Terry Dykstra wrote: > > You can use InfoMaker (Sybase). It has a so called pipeline utility > > which allws you to move data between ant 2 databases. Very easy to use. > > > Terry Dykstra > > My favorite utility for complex cases is still Perl although I am reading > a Python book and I like what I read. If only there was a comprehensive > Python archive network, with tons of modules and applications available, > Python would immediately take over the world by storm. PyPI is not nearly > as good - yet. > > --http://mgogala.byethost5.com
From: Steve Howard on 14 Jun 2010 07:26 On Jun 13, 5:14 pm, Mladen Gogala <gogala.mla...(a)gmail.com> wrote: > > My favorite utility for complex cases is still Perl although I am reading > a Python book and I like what I read. If only there was a comprehensive > Python archive network, with tons of modules and applications available, > Python would immediately take over the world by storm. PyPI is not nearly > as good - yet. > > --http://mgogala.byethost5.com I *love* python. 98% of all tools/scripts I write use python, which has been the case for only about 18 months or so. I have not had a problem finding anything I need in terms of a class library, although there is nothing equivalent to cpan.
From: Mladen Gogala on 14 Jun 2010 10:09 On Mon, 14 Jun 2010 04:26:45 -0700, Steve Howard wrote: > I *love* python. 98% of all tools/scripts I write use python, which has > been the case for only about 18 months or so. I fell in love in python when I realized that I can use files as iterators like this: myfile=open(pathname,"r") for line in myfile: line=line.rstrip('\n') # This is "chomp". Slightly longer than the # Perl version, but not too bad. .... do some processing This is beautiful. No unclear "diamond notation", no "while" loops, just a simple and very logical iteration through the file lines. I am somewhat biased because I have been working with Perl since 1996 and was also using Perl 4, with oraperl, of course. I feel completely comfortable and at home with Perl and it will be a long time before I have the same feeling with Python. Unfortunately, Perl is huge, there are many elements to learn, Perl is somewhat illogical (count of the elements in an array @arr is obtained through the conversion to scalar context (scalar(@arr)) and has some strange idioms, like $a ||= 0; which really means if (!defined($a)) { $a=0; }. CPAN is also the Perl's biggest problem because any new version will not be fully adopted until the modules people are using are ported to the next version. With the Perl's atrocious object model, all the modules will inevitably be broken in Perl6. I am afraid that Perl 6 will be syntactically even larger than Perl 5, which means that it will require a significant learning curve, just like the transition from Perl4 --> Perl5. I must say that at this point, I consider learning Python a better alternative to remaining with the huge and slowly developing Perl which has a very strange object model indeed. -- http://mgogala.byethost5.com
From: joel garry on 14 Jun 2010 13:03
On Jun 13, 9:20 pm, "if_inves...(a)yahoo.com" <basis_consult...(a)hotmail.com> wrote: > Hi, > > Thanks for the replies. Some people asked me for more details about > the > issue. > > There is an MSSQL 9.00 database about 100GB in size that we need to > get a copy of; To be more accurate, we need to get hold of its major > schema, and create an Oracle database (Hoping that it will be 10g) > with > the same schem, having the same data. (The entire database consists of > only > that schema and whatever the default MSSQL schemas are). > > I am not sure if the source DB has unique MSSQL features; I don't know > MSSQL, and I am not sure sure what to check. Ah, be careful about tripping over terminology. What Oracle calls a schema, MS calls a database. You'll need to be conscious of this with some tools. Also see http://www.oracle.com/technology/tech/migration/index.html (which is a slightly longer version of what Mladen said in two words). Steve has a good point about using existing expertise. Sometimes that should control the tools decision, sometimes new tools should be investigated. As always, it depends. I'm always fighting with having to let users use Excel as a data loading tool. jg -- @home.com is bogus. http://www.nytimes.com/2010/06/07/technology/07brain.html?hp |