Re:SQLite – The “server-process-edition” branch
by nbevans on Wednesday August 15, 2018 @04:44PM
(#17769595)
... as long as those threads don't touch the same page as each other. Page locks <> row locks.
Re:SQLite – The “server-process-edition” branch
by blattimwind on Wednesday August 15, 2018 @06:53PM
(#17770453)
Depends on the width of your rows and the relation of access patterns to physical row allocation.
Re:SQLite – The “server-process-edition” branch
by hd4 on Wednesday August 15, 2018 @02:02PM
(#17768256)
For some reason, I had the idea that this was always the case with SQLite, at least for the last 10 years or so.
Re:SQLite – The “server-process-edition” branch
by blattimwind on Wednesday August 15, 2018 @02:03PM
(#17768267)
In WAL mode, readers don't block writers and vice versa, but there can still only be one writer active at any given time.
No.
> (b) Does this mean that sqlite will now have the capability to do concurrent reads/writes, but that it is upto users to implement a process that can take advantage of this and create something like (a)?
Yes... this allows a multi-threaded application for example to have multiple readers/writers without blocking each other.