Re: [GENERAL] Any justification for sequence table vs. native sequences?

2009-08-18 Thread Tom Lane
Bill Moran writes: > And yes, it's pretty much guaranteed to be slower than built in sequences, > with > blocking when multiple threads want a sequence all at the same time. It's also going to create a vacuum bottleneck unless the insert rate is quite low, because each ID assignment will create

Re: [GENERAL] Any justification for sequence table vs. native sequences?

2009-08-18 Thread Bill Moran
Doug Gorley wrote: > > I just stumbled across this table in a database > developed by a collegue: > > > field_name | next_value | lock > +-+ > id_alert| 500010 | FREE > id_page | 500087 | FREE > id_group| 500021 | FREE > > > These "i

Re: [GENERAL] Any justification for sequence table vs. native sequences?

2009-08-18 Thread Stuart McGraw
On 08/18/2009 01:14 PM, Doug Gorley wrote: I just stumbled across this table in a database developed by a collegue: field_name | next_value | lock +-+ id_alert| 500010 | FREE id_page | 500087 | FREE id_group| 500021 | FREE These "id

[GENERAL] Any justification for sequence table vs. native sequences?

2009-08-18 Thread Doug Gorley
I just stumbled across this table in a database developed by a collegue: field_name | next_value | lock +-+ id_alert| 500010 | FREE id_page | 500087 | FREE id_group| 500021 | FREE These "id_" fields correspond to the primary keys on