I have a RouterBOARD 133 with RouterOS 2.9.42, configured as a transparent traffic shaper (on LAN, no Wifi), shaping a 10MB line, about 200 users, CPU load about 20% with some spikes up to 40%.
The RB freezes every couple of minutes, I had to set the Watchdog to ping (to the gateway), then it reboots after the freeze, but this is no real solution.
The log shows only: "watchdog cannot ping address X.X.X.X, rebooting"
What could be the problem?
- too much traffic
- too many users
the CPU load seems to be OK
P.S. The 10MB line and 200 users is a stress test, the RB is planned for a 1MB line, 20-30 users, but will it be stable?
Seems not to be a bad Board, I have tried a second one with the same result 
I have also tried the RB on a small LAN - 256k line, 6 users, no freeze till now.
A bad config can cause freezes with higher loads? OK, I'll try to play with the config ...
I've instaled RouterOS demo licence on a PC, loaded the config from the RB and it work flawless, yet ...
So the config seems to be OK.
Shall I try the beta RouterOS on the RB or is it meaningless and the RBs are simply bad?
I have also tried the RB on a small LAN - 256k line, 6 users, no freeze till now.
A bad config can cause freezes with higher loads? OK, I'll try to play with the config ...
I've instaled RouterOS demo licence on a PC, loaded the config from the RB and it work flawless, yet ...
So the config seems to be OK.
Shall I try the beta RouterOS on the RB or is it meaningless and the RBs are simply bad?
thanksI tried the serial console (Watchdog ping disabled), when the RB freezes (after cca 30 min of uptime), the ethernet interfaces (/interface ethernet monitor) have no link and autonegotiation is disabled, the rest of the RB seems to work normally, nothing in the log.
After a couple of minutes the link is back, autonegotiation still disabled, speed is down from 100MB to 10 MB, network connection to the RB not working.
I rebooted the RB via the serial console, it took cca 5 minutes to the next freeze. This time also the serial console was frozen, after cca one hour the console was "unfrozen", the rest was the same, nothing in the log, no network connection.
Maybe there is some more detailed log, where I can figure out, what is happening with the RB?
After a couple of minutes the link is back, autonegotiation still disabled, speed is down from 100MB to 10 MB, network connection to the RB not working.
I rebooted the RB via the serial console, it took cca 5 minutes to the next freeze. This time also the serial console was frozen, after cca one hour the console was "unfrozen", the rest was the same, nothing in the log, no network connection.
Maybe there is some more detailed log, where I can figure out, what is happening with the RB?
-
-
evolutionise
just joined
- Posts: 8
- Joined:
I've tried 2 different power supplies, no (significant) voltage drop on load.
I've also tried to connect the RB between two PC and made a 100% RB CPU load by file copying - the RB was stable, so I don't think it is a power supply problem.
I've also tried several CAT5E cables and the RB still freezes ...
I've also tried to connect the RB between two PC and made a 100% RB CPU load by file copying - the RB was stable, so I don't think it is a power supply problem.
I've also tried several CAT5E cables and the RB still freezes ...
-
-
arizonajon
just joined
- Posts: 12
- Joined:
Same Problem, but on an RB112
Sounds like the problem I'm facing on a link using RB112 boards. Mine are at a remote site, so I cannot see the LEDs.
After anywhere from minutes to hours, the RB appears to stop listening on eth1 and will not respond to pings. Does not appear to be too much correlated to activity on the wlan1 port or the eth1 port.
Power 13vdc, 1A regulated) is supplied via the coaxial input, not PoE. Power appears clean and reliable. Seems that "sometimes" it spontaneously recovers after an indefinite period, but on the other hand I've had to call someone to cycle power.
RB eth1 is 192.168.1.26 on a local subnet 192.168.1.0/24, there's a Linksys WRT54G AP/router/gateway at 192.168.1.1, and my remote access computer is via ethernet at 192.168.1.80. I surf in via ultraVNC to that machine, and am generally able to control the RB remotely. The RB talks via WDS to an even more remote 192.168.2.0/24 subnet.
No serial access from my location, will have to travel 40 miles to get to the DB9, which may be the next thing I have to do...
Cheers - Jon
After anywhere from minutes to hours, the RB appears to stop listening on eth1 and will not respond to pings. Does not appear to be too much correlated to activity on the wlan1 port or the eth1 port.
Power 13vdc, 1A regulated) is supplied via the coaxial input, not PoE. Power appears clean and reliable. Seems that "sometimes" it spontaneously recovers after an indefinite period, but on the other hand I've had to call someone to cycle power.
RB eth1 is 192.168.1.26 on a local subnet 192.168.1.0/24, there's a Linksys WRT54G AP/router/gateway at 192.168.1.1, and my remote access computer is via ethernet at 192.168.1.80. I surf in via ultraVNC to that machine, and am generally able to control the RB remotely. The RB talks via WDS to an even more remote 192.168.2.0/24 subnet.
No serial access from my location, will have to travel 40 miles to get to the DB9, which may be the next thing I have to do...
Cheers - Jon
-
-
jandafields
Forum Guru
- Posts: 1515
- Joined:
I have the same problem where one of my RB133 boards will randomly lockup. The only solution is a power cycle.
Although it crashes randomly, I can force it to crash by doing this:
http://forum.mikrotik.com/viewtopic.php?t=15070
Although it crashes randomly, I can force it to crash by doing this:
http://forum.mikrotik.com/viewtopic.php?t=15070
Who is online
Users browsing this forum: No registered users and 5 guests