get next implementation question

Hi

I'm using net-snmp-5.8 agent code to implement my own MIB table.
My table will have 10000+ rows stored in external database.

I used  mib2c.iterate.conf to create skeleton code for my table, which
generated two functions:
xxx_get_first_data_point and xxx_get_next_data_point. Below is my current
implementation:

1. I created a linked list to store the 10000+ records;
2. In xxx_get_first_data_point(), I set "my_loop_context" to the header of
the linked list, and call the xxx_get_next_data_point();
3. In xxx_get_next_data_point(), set "my_loop_context" to next item of the
linked list, and return the "my_context_data", or NULL if the
"my_loop_context" is NULL.

It worked, however, it seems not too efficient:  each time  when call the
xxx_get_first_data_point(), it will loop from the header of the linked list;

My question is:
1. Can I know the MIB table index from xxx_get_first_data_point() or
xxx_get_next_data_point(), so that I needn't loop from the header of the
linked list?
2. Can I not to create linked list firstly, just query from data base every
time when there is snmp get next request comes in?

Thanks
Spark

Thanks
Spark

_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders