Routing table is too small in 382.x

[kaitzu]

Hi,

I have dual wan configuration with load balancing. It screws the internet connection normally, but I use it with routing table. My home setup was
x.x.x.2 - x.x.x.63 addresses were reserved to fixed IP devices
x.x.x.64 - x.x.x.127 to dhcp
Even addresses are routed to wan1 and odd to wan2. I had exactly 128 entries in routing table.
This has been working at least in 380.x, but I noticed today (382.x) that all addresses were routed to wan1. I tried to fix it without success. Dropped entries down to 64, but no luck. I'm using gui.

I decided to reset to factory settings. Then I entered 64 entries (as the new fw limit for routing table entries has changed from 128 to 64) to routing table, but device didn't save them. Just showed up an "success" popup and returned to empty routing table. Then I tried to add routes small bunch at time and found out maximum routes I can enter to my device is 51. Checked log and trying to add even one gives error:
"Dec 7 22:31:32 httpd: nvram_check fail: nvram wans_routing_rulelist over length (1028 > 1024)"

I'd like to have more entries in routing table. Is there any configuration option for that?
My router is RT-AC88U.

Kai

 

[RMerlin]

I'd like to have more entries in routing table. Is there any configuration option for that?
My router is RT-AC88U.

It's a build-time parameter.

I can try increasing it to 2048, but I'm not entirely sure the rest of the firmware will properly accept it (and the RT-AC86U will certainly not allow it, that model is hardcoded by Broadcom to a set limit of 1000 per nvram settings)

 

[kaitzu]

It's a build-time parameter.

I can try increasing it to 2048, but I'm not entirely sure the rest of the firmware will properly accept it (and the RT-AC86U will certainly not allow it, that model is hardcoded by Broadcom to a set limit of 1000 per nvram settings)

If you try it in next betas I'll give it a try

Lähetetty minun SM-N950F laitteesta Tapatalkilla

 

[sfx2000]

Another way of looking at things if one wants to LAN ranges to WAN interfaces. Instead of doing like OP, where even goes to WAN1 and odd to WAN2, bind ranges instead - and this makes things a bit easier...

Static Range - x.x.x.2/26
LAN block 1 - x.x.x.64/26
LAN block 2 - x.x.x.128/26

Normally though -- in Dual WAN, esp. if both WAN's are from the same ISP, is one has the HA interface, and not really worry about which LAN is bound with which WAN... keeping in mind that dual WAN is more about uptime and failover, rather than binding LAN clients to a specific WAN

y.y.y.32/30 (WAN1) -----> HA ----- x.x.x.0/24 (LAN) - maybe do some /26 blocks for VLAN
y.y.y.42/30 (WAN2) --/

And then one can get creative with VLAN's - VLAN101 could be one range, VLAN102 could be another, and so forth.

 

[kaitzu]

Another way of looking at things if one wants to LAN ranges to WAN interfaces. Instead of doing like OP, where even goes to WAN1 and odd to WAN2, bind ranges instead - and this makes things a bit easier...

Static Range - x.x.x.2/26
LAN block 1 - x.x.x.64/26
LAN block 2 - x.x.x.128/26

Normally though -- in Dual WAN, esp. if both WAN's are from the same ISP, is one has the HA interface, and not really worry about which LAN is bound with which WAN... keeping in mind that dual WAN is more about uptime and failover, rather than binding LAN clients to a specific WAN

y.y.y.32/30 (WAN1) -----> HA ----- x.x.x.0/24 (LAN) - maybe do some /26 blocks for VLAN
y.y.y.42/30 (WAN2) --/

And then one can get creative with VLAN's - VLAN101 could be one range, VLAN102 could be another, and so forth.

I do have both WAN from same ISP. The "free load balancing" problem might be that both are 4G modems. Still I've bought public dynamic IP's to both and they're not behind "thoussands of NAT". My house is in the place where copper is more than 3km (~10kfeet) long and xDSL just can't do it fast enough. I'll give a try to your simplified routing rules, thanks!

 

[sfx2000]

And to explain the ranges suggested... VLAN's work well here, port based perhaps if working with the primary router - maybe define the groups there, and managed switch behind it - and one can still do resource sharing...

$ ipcalc 192.168.1.2/26
Address:   192.168.1.2          11000000.10101000.00000001.00 000010
Netmask:   255.255.255.192 = 26 11111111.11111111.11111111.11 000000
Wildcard:  0.0.0.63             00000000.00000000.00000000.00 111111
=>
Network:   192.168.1.0/26       11000000.10101000.00000001.00 000000
HostMin:   192.168.1.1          11000000.10101000.00000001.00 000001
HostMax:   192.168.1.62         11000000.10101000.00000001.00 111110
Broadcast: 192.168.1.63         11000000.10101000.00000001.00 111111
Hosts/Net: 62                    Class C, Private Internet

$ ipcalc 192.168.1.64/26
Address:   192.168.1.64         11000000.10101000.00000001.01 000000
Netmask:   255.255.255.192 = 26 11111111.11111111.11111111.11 000000
Wildcard:  0.0.0.63             00000000.00000000.00000000.00 111111
=>
Network:   192.168.1.64/26      11000000.10101000.00000001.01 000000
HostMin:   192.168.1.65         11000000.10101000.00000001.01 000001
HostMax:   192.168.1.126        11000000.10101000.00000001.01 111110
Broadcast: 192.168.1.127        11000000.10101000.00000001.01 111111
Hosts/Net: 62                    Class C, Private Internet

$ ipcalc 192.168.1.128/26
Address:   192.168.1.128        11000000.10101000.00000001.10 000000
Netmask:   255.255.255.192 = 26 11111111.11111111.11111111.11 000000
Wildcard:  0.0.0.63             00000000.00000000.00000000.00 111111
=>
Network:   192.168.1.128/26     11000000.10101000.00000001.10 000000
HostMin:   192.168.1.129        11000000.10101000.00000001.10 000001
HostMax:   192.168.1.190        11000000.10101000.00000001.10 111110
Broadcast: 192.168.1.191        11000000.10101000.00000001.10 111111
Hosts/Net: 62                    Class C, Private Internet