proper OID for netsnmp_register_cache_handler

Hello,

I've been experimenting with the caching (netsnmp 5.3.1-14.el5 yeah
RedHat) and I've noticed something that I do not understand. First of
all here are the details:

mod ::= t 1.3.6.1.4.1.1139.16.10  -- top level module OID 
sw ::== { modMIB 1 } //SW_OID
swTable ::= { sw 1} // SWTABLE_OID

I'm basically just filling a table of RPM packages in swTable.

Here is my init routine:

     55 void init_software(void)
     56 {
     57     static oid swTable_oid[] = { SWTABLE_OID };
     58     static oid sw_oid[] = { SW_OID };
     59     netsnmp_table_registration_info *table;
     60     netsnmp_handler_registration *handler;
     61     netsnmp_iterator_info *iter;
     62     int rc = 0;
     63
     64
     65     table =
SNMP_MALLOC_TYPEDEF(netsnmp_table_registration_info);
     66     iter = SNMP_MALLOC_TYPEDEF(netsnmp_iterator_info);
     67
     68     handler = netsnmp_create_handler_registration("swTable",
     69
swTable_handler,
     70                                                   swTable_oid,
     71
OID_LENGTH(swTable_oid),
     72
HANDLER_CAN_RONLY);
     73     if (!handler || !table || !iter)
     74     {
     75         snmp_log(LOG_ERR, "malloc failed in init_software");
     76         return;
     77     }
     78
     79     netsnmp_table_helper_add_indexes(table, ASN_INTEGER, 0);
     80     table->min_column = SWTABLE_MIN_COLUMN;
     81     table->max_column = SWTABLE_MAX_COLUMN;
     82     iter->get_first_data_point = swTable_get_first_data_point;
     83     iter->get_next_data_point = swTable_get_next_data_point;
     84     iter->table_reginfo = table;
     85     netsnmp_register_table_iterator(handler, iter);
     86
     87
     88     rc =
netsnmp_register_cache_handler(netsnmp_create_handler_registration
     89                                         ("swTableCache", NULL,
     90                                          sw_oid,
OID_LENGTH(sw_oid), 0), // fails if I use swTable_oid
     91                                         SWTABLE_CACHE_TIMEOUT,
     92                                         swTable_load,
swTable_free);
     93
     94     DEBUGMSGTL(("sdp", "netsnmp_register... = %d\n", rc));
     95 }

The code above does work.

What I have noticed is that if I call netsnmp_register_cache_handler
with the OID for the table (swTable_oid) the function returns -1 and the
snmp walk returns nothing. If I use an OID one level up, either sw_oid
(.16.10.1) or mod_oid (.16.10) the routine returns 0 and the snmp walk
works. I don't understand why using the same OID for the two calls
(lines 68, 88) fail. My intent is to build a module which has a number
of caches, each with different tables and load routines, that's why I
cannot use the higher level OIDs so I thought that I can just use the
same OID as the table. Is the cause of the problem due to the two calls
to netsnmp_create_handler_registration for the same OID? I also tried
using the same handler from the first call to
netsnmp_create_handler_registration (line 68) and that also did not
work.  Can someone please explain. Thanks.

---
Don Khan



------------------------------------------------------------------------------
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Net-snmp-users mailing list
[email protected]
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users