You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a bit of a dual question:
1st. Is there a better place I should be asking these non-issue clarification questions that the docs alone were unable to answer for me?
2nd: What are a some best practices for using ZODB in a ZEO configuration with a handful of users that will be operating(r/w) mostly in their own BTree persistences? specifically I am trying to employee best practices when trying to read from the database such that I don't overwrite work by attempting to refresh the current view of the 'live' data.
Any assistance would be much appreciated as I am very new to ZODB and have chosen it as we need a near embedded database instance that can be easily traversed to find groupings of applicable data where the structure can be vastly different between the entries beyond core 'queryable' tags that are used in the keys of the BTree structures mostly.
The text was updated successfully, but these errors were encountered:
...
2nd: What are a some best practices for using ZODB in a ZEO configuration with a handful of users that will be operating(r/w) mostly in their own Bree persistence's? specifically I am trying to employee best practices when trying to read from the database such that I don't overwrite work by attempting to refresh the current view of the 'live' data.
As you (hopefully) know, the ZODB implements (for the typical
storages) snapshot isolation: i.e. it delivers object
state valid at a snapshot time. The snapshot is
defined as the latest locally known transaction when the
ZODB connection was opened or the current transaction begun, respectively.
If your system is distributed, then the local information
may be incomplete: there might be transactions which
have been successfully completed at the open/transaction start time
which the local system does not yet know about.
If your system components interact with one another, they
may observe inconsistencies due to this behavior.
ZEO's `ClientStorage` has the option `server_sync`.
If you set it to `True`, then at transaction start
the server is called ensuring that the local system
knows about all transactions completed before the transaction start time.
@navitux is major ZODB expert regarding concurrency.
You may want to search for his comments in this repository.
He usually publishes detailed reports; your search may deliver
corresponding entry points.
I have a bit of a dual question:
1st. Is there a better place I should be asking these non-issue clarification questions that the docs alone were unable to answer for me?
2nd: What are a some best practices for using ZODB in a ZEO configuration with a handful of users that will be operating(r/w) mostly in their own BTree persistences? specifically I am trying to employee best practices when trying to read from the database such that I don't overwrite work by attempting to refresh the current view of the 'live' data.
Any assistance would be much appreciated as I am very new to ZODB and have chosen it as we need a near embedded database instance that can be easily traversed to find groupings of applicable data where the structure can be vastly different between the entries beyond core 'queryable' tags that are used in the keys of the BTree structures mostly.
The text was updated successfully, but these errors were encountered: