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
The value of time_threshold = 100 often brings hardly analyzed results in case of many 'lightweight' queries.
For instance, most of all requests of replication system (in case of proper configuration) will not be logged with threshold more than 30-40 ms.
I think that if someone turned trace on, all enabled events should be logged until time threshold is explicitly set to some non-zero value.
The text was updated successfully, but these errors were encountered:
Your request assumes that the Trace functionality is expected to be used as an "audit" of all database operations. Although this is one possible use of the function, it is an exceptional mode for most installs.
Someone looking to use the Trace should inform themselves about how it should be configured. For them to "assume" that all events will be logged is an un-informed method of operation.
Just as you have suggested that the threshold should be 0, it would be reasonable for someone to complain that the default setting generates too many entries/ too much detail... so, no default value will satisfy all users. Personally, I would be looking for a default of 5 sec since I would be looking for long running queries.
Submitted by: @pavel-zotov
The value of time_threshold = 100 often brings hardly analyzed results in case of many 'lightweight' queries.
For instance, most of all requests of replication system (in case of proper configuration) will not be logged with threshold more than 30-40 ms.
I think that if someone turned trace on, all enabled events should be logged until time threshold is explicitly set to some non-zero value.
The text was updated successfully, but these errors were encountered: