Activity overview
Latest activity by MeshMan
I don't think it's network latency as the comparison is done purely on local across 2 DB instances when I'm testing.
Your SQL Compare tool can compare any two databases in about 30 seconds, yet a tiny wrapper over the SDK takes 15-20 mins easy. Perhaps it's something we're not doing right, yet there isn't much code. I see it simply building a differences array, selecting them all, and then executing the work through a BlockExecutor...
Sorry to digress. So, I'll give the two pass approach a try but meanwhile, need to look at sorting out the time its taking.
Thanks,
Martin. / comments
I don't think it's network latency as the comparison is done purely on local across 2 DB instances when I'm testing.
Your SQL Compare tool can compare any two databases in about 30 seconds, yet a t...
Hi Michelle,
You've hit the nail on the head there. I imagined leaving the database in an invalid state which then I would populate the relevant tables and then re-enable all the foreign key constraints now that the values would be valid in the new column.
However, I don't think the synch library allows creating the foreign key constraints but leaving them disabled?
I loved your idea about having a temporary table for moving the data into, then truncating the changing table. The only problem with this is that any depending tables would also need their data moving or constraints disabling. That could get really messy?
How about a 2 pass approach where I ask the synch to ignore keys, then populate my tables with data, then run a another synch to then include the keys. What worries me about that though is that the compare already takes about 20 minutes to perform, I would hate to think would get longer by another substantial amount of time.
The database in question has:
462 tables.
483 views.
4,682 procedures.
3,946 constraints.
Any more thoughts would be appreciated.
Regards,
Martin / comments
Hi Michelle,
You've hit the nail on the head there. I imagined leaving the database in an invalid state which then I would populate the relevant tables and then re-enable all the foreign key constr...
Can I get anymore help with this as I'm still not aware of any way of getting around this without disabling foreign key constraints which would be useless as I need them all comparing/synching etc.
So the rules are, never allow null values on a column, inserting a foreign key column that is referencing a new table that has no data - what's the best way of handling this? / comments
Can I get anymore help with this as I'm still not aware of any way of getting around this without disabling foreign key constraints which would be useless as I need them all comparing/synching etc....
That's not really going to help me then as when new tables are created or existing ones changed, the foreign keys still need to be synch'd.
I'm thinking that I will just have to make the specific table change myself with a pre-run script, but that goes against the use of the engine. :? / comments
That's not really going to help me then as when new tables are created or existing ones changed, the foreign keys still need to be synch'd.
I'm thinking that I will just have to make the specific t...
New Table Column - Foreign Key Constraint
Hi,
Can someone please advise me how to get around the following problem?
The database I'm working with must conform to a column having a default value and no nulls are allowed. So, foreign key col...