gnrc_rpl: introduce control plane traffic statistics#5381
Conversation
fbafa77 to
b6c7133
Compare
|
I guess its not waiting anymore. |
| #endif | ||
| /* a leaf node should only react to unicast DIS */ | ||
| && (gnrc_rpl_instances[i].dodag.node_status != GNRC_RPL_LEAF_NODE)) { | ||
| if(gnrc_rpl_instances[i].dodag.node_status != GNRC_RPL_LEAF_NODE) { |
|
Don't know if it makes sense, but how about also counting errors, e.g. invalid DIOs/DAOs? |
ff426cf to
7403fb4
Compare
|
@BytesGalore rebased onto master and did a little overhaul. The statistics aren't bound to the instances anymore, because they would get lost otherwise when a node leaves the dodag. Your idea about counting errors is also nice, but let's concentrate on this for another PR (: getting too big again. You can test by compiling |
| uint32_t dao_ack_tx_ucast_bytes; /**< unicast dao_ack sent in bytes */ | ||
| uint32_t dao_ack_tx_mcast_count; /**< multicast dao_ack sent in packets */ | ||
| uint32_t dao_ack_tx_mcast_bytes; /**< multicast dao_ack sent in bytes*/ | ||
| } netstats_rpl_t; |
There was a problem hiding this comment.
Wouldn't it make maybe more sense to put this into some RPL specific header?
There was a problem hiding this comment.
would it make sense to introduce a counter type to easy switching smaller/larger nummeric types, e.g. uint16_t, like:
typedef uint16_t netstat_rpl_cnt_t;
typedef struct {
/* DIO */
netstat_rpl_cnt_t dio_rx_ucast_count; /**< unicast dio received in packets */
...
}netstats_rpl_t;?
|
@cgundogan do you think you can address the comments during the next week? |
|
Well, let's see for the next release. |
7403fb4 to
6989ea2
Compare
|
I moved the rpl related netstats struct into a separate file as suggested by @OlegHahm |
|
Somehow the comment of @BytesGalore regarding a configurable struct member type got lost (commented on commit? (; ). But we can also do this as a follow-up, what do you think @BytesGalore ? |
|
@cgundogan you mean 8f1d438#r65735118 : typedef uint16_t netstat_rpl_cnt_t;
typedef struct {
/* DIO */
netstat_rpl_cnt_t dio_rx_ucast_count; /**< unicast dio received in packets */
...
}netstats_rpl_t;no objection to put/discuss it in a followup PR |
|
I would say please squash then ACK from my side |
|
@cgundogan, ping! |
|
@cgundogan I still would say please squash then ACK from my side |
43b83f1 to
16bfcdf
Compare
|
@BytesGalore squashed |
16bfcdf to
cfb487f
Compare
Provide statistics about RX/TX DIOs, DISs, DAOs, DAO-ACKs.
based on #5178