blob: 3d2117df537cdbf00fff4fcee70c6e958c2c3860 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
|
Sparse notes on major existing problems/plans
* SMI issues remained include:
1) multiple OID registration should be posible for
customizing Managed Objects classes
2) in MIBVIEW object model: MIB module name, MIB object
name and object instance should be clearly separated.
Dedicated, simple API for table cells mangling should be
provided.
* Walking over VACM shadowed OIDs is painfully slow
* Notification filtering (RFC3413) not implemented.
* MsgAndPduDispatcher should use TransportDispatcher jobs interface
to handle SNMPv3 handshake
* Implement 'oneliner' API for v1arch, possibly compatible with v3arch API
* Implement prettyIn value format parser at TC
* Disallow SNMP access to USM objects
* Maybe MibInstrumentationController should be async what
implies callback function interface.
* Re-work ACL to get rid of linear search there
* Implement partial context name match at isAccessAllowed()
* Sort out whether pyasn1 objects should probably have either clone() or
subtype() methods.
* Review PDU components coercion code at proxy/rfc2576
* TextualConvention subclassing order now depends upon DisplayHint
clause presense in MIB definition (for useful prettyIn/Out()). Maybe
this should be re-worked.
* exclude SNMP-COMMUNITY-MIB from view for better security
* re-design something to get rid of an ugly hack coercing MIB values into
SNMP message header as it has incompatible constraints set
* Implement OID prefix to MIB module mapping then dynamic MIB modules
loading on OID to MIB symbol resolution
* Implement TCP transport support
* Rework UNIX sockets transport to make it STREAM-type.
* The get-first/get-last operations of pysnmp.smi.rfc1902.ObjectIdentity()
may need a means to deal only with specific node types.
* redesign proto.errind.ErrorIndication
* move LCD and var-bind management routines from Async* classes
* make oneliner helper routines more general, not bound to asyncore-based
implementation
* improve oneliner modules layout to make it possible to create more
transport-specific oneliner interfaces
|