partition authdb

[chaen]

No description provided.

[fstagni]

What if the DB was already created? Should the partitioning be done by hand?

[chrisburr]

What if the DB was already created? Should the partitioning be done by hand?

DiracX should do it for them. If they already have a large AuthDB (which should only be LHCb) then some more care is needed.

[chaen]

What if the DB was already created? Should the partitioning be done by hand?

DiracX should do it for them. If they already have a large AuthDB (which should only be LHCb) then some more care is needed.

Not if there is already data in the table. In this case, we print the command to execute.

[fstagni]

Not if there is already data in the table. In this case, we print the command to execute.

Yes, my question was if the DB is already created AND there's data in it (does is matter how much?).

"we print the command to execute.". Who/what is "we" ?

[chaen]

if you have 2 billion rows like LHCb, you want to do it in a bit smarter way than just running the alter schema command. Fortunately, I think it is not the case of anyone :-)
The init-sql job prints that in its output.