You are here

OLSRD Scalability

3 posts / 0 new
Last post
AE6XE
AE6XE's picture
OLSRD Scalability

An olsrd scalability note:   Today we had about 117 nodes on the tunneled meshes.   I did a quick capture of olsrd packets.   From AE6XE-10 (MPR) node olsrd is sending at least 1 packet every second, and frequently 3 packets every second.  The HNA and MID advertisements (quick observation) looks like 75% of the information.  Most packets are pushing 1500 in size.     This link was 1.0 MBit/s both directions--according to iw station dump.  

 

KG6JEI
Links will always be 1.0mbit

Links will always be 1.0mbit RF speed for broadcasts (iw station dump is about the RF not the data) due to how 802.11 all broadcast packets have to go out at the slowest possible network speed so that the slowest device may receive them. 

The only way to change the minimum RF speed this would be to implement a minimum data rate function.

1500 size makes sense as the program is designed to combine as many packets as it can.

 There may be a need to extend out and space tunnel packets in some manner however I'm not sure that can be done off hand. 

3 packets per second would be only around a max of 4.5 kb/s (plus RF frame overhead) however this does illustrate why I am always so paranoid about any suggestion to add extra data to OLSRD.

Admin note:

Topic moved to general discussion

KD0LIX
Do we have 802.11e's frame
Do we have 802.11e's frame aggregation support? If so, it might make sense to go with a slightly smaller packet size, as you could fit two in a 2304-byte WiFi MTU. But, I suspect for broadcast packets each node would transmit with least-common-denominator headers.

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer