Beta 9 on 133c kills the board, can’t even mac terminal in.
I tried on 4 different boards, all with minimal packages.
I just did a rb153 with b9 and it’s rebooting after 5-30 seconds after becoming pingable.
Disabled a couple packages and same results. Will have to swap rb to fix it.
b8 wouldn’t let me telnet/ssh into it, just winbox, though it passed traffic reliably, calea didn’t work on it. No wireless, just 3 ethernets.
Just netinstalled a 133c system with only wireless package and system package of course.
The system runs smoothly until I install the NTP package.
/Paul
i installed beta9 on 133C it was unresponsive
did system reset (it had ip address et up anyway)
logged in through console added ip address, and everything is ok
loaded packages:
[admin@MikroTik] > system package print detail
Flags: X - disabled
0 name="advanced-tools" version="3.0beta9" build-time=may/31/2007 16:20:03 scheduled=""
1 name="routerboard" version="3.0beta9" build-time=may/31/2007 16:21:15 scheduled=""
2 name="ppp" version="3.0beta9" build-time=may/31/2007 16:20:16 scheduled=">
3 name="security" version="3.0beta9" build-time=may/31/2007 16:20:08 scheduled=""
4 name="system" version="3.0beta9" build-time=may/31/2007 16:19:56 scheduled=""
5 name="wireless" version="3.0beta9" build-time=may/31/2007 16:20:27 scheduled=""
6 name="routeros-rb500" version="3.0beta9" build-time=may/31/2007 15:22:04 scheduled=""
7 X name="routing" version="3.0beta9" build-time=may/31/2007 16:20:18 scheduled=""
8 name="ntp" version="3.0beta9" build-time=may/31/2007 16:20:39 scheduled=">
9 name="dhcp" version="3.0beta9" build-time=may/31/2007 16:20:10 scheduled=""
10 name="option" version="3.0beta9" build-time=may/31/2007 16:21:43 scheduled=""
[admin@MikroTik] >
i have something about 4MB free memory at the moment
Same problem here with an RB112, just upgraded from 2.9.42 to 3.0beta9 boord hangs after reboot.
No mac telent nothing, can’t get it to see netinstall…writing this while waiting for serial install back to 42.
Is this a known issue?
I’ve seen similar sounding issues with beta 8 and 9, but won’t know details until I get back to the office monday, I’m still in florida after MUM…
We got B9 to work on 133c boards with atheros based card, but the AP->CPE/Beta9/133c throughput is no good.
The throughput grows and grows then crashes and sometimes the board reboots. Looks like it might be a buffer overrun? Funny thing is I have seen a Ubiquiti LS5 exhibit the same behavior. Beta6 didn’t have this problem. When you go from beta9 back to beta6 the wireless card comes back up disabled and back to defaults… Not so fun…
Robert
I was able to bring this box into the office today and netinstall it. Seems to work fine with b9 after that.
the only boards that could have problems is small CPE with 16MB of ram
but after some tinkering my RB133c and RB112 is working fine
And what was the tinkering you did?
i killed a 112 today with b9. It reboots, reboots, reboots. Sometimes it will give me a terminal login banner but never lets me login as its already rebooting again. It would be nice if that button on the board would allow a PXE boot - no null modem cable to be found : (
Sam
beta9 does not work on this 112 board… 1.13 and 2.8 firmware. rolled back to 2.9.42 and its working as expected. something is weird with beta9 and these boards i believe. they boot partially but almost never double beep and complete startup. Tried clean install with netinstall, as well as xmodem’ed a new 2.8 firmware.
Sam
as beta is beta - i did /system reset-configuration, then when the system came up it was responsive (connected through the console)
then added ip address route, thats all
if upgrades with this little configuration - system was quite unresponsive.
I just tried an upgrade to a RB133C with what appeared initially to be a bricked outcome…
The netinstall went OK with the full package.
Unit rebooted but no console login or winbox connectivity
I did a config reset from the bootloader, exited & waited followed by a cold boot
Got a console after this but it still took a while ( 3..4 minutes?? ) after the banner before I got a commandline
Still no Winbox at this stage
Gave the unit an IP address
Now Winbox connected & downloaded plugins before giving me full access
rebooted and had the similar console signon delay but this time no winbox connectivity. Could see the MAC but got timeout error messages when trying to connect. Unit pinged OK.
Will try a smaller/ leaner install package next
Cheers
Don
I tried just loading the system & PPP packages. This time the Rb133C appeared stable in this configuration.
I then upgraded a second unit and ran Bandwidth test over the async PPP link between the two RB133C’s.
Nice solid 44kbps TCP running bidirectional with a 512 byte MTU. This cured the problem we were having with PPP over async on the older 2.9.43. e.g. good with UDP but problems with TCP.
When time permits I’ll add some more packages & retest.
Cheers
Don
One of the units locked up when running packet sniff. After reboot it allowed access via winbox & I noticed that it had erased all the routes table & refused terminal access.
Tried cold boot and configuration reset via the boot menu.
Now I have terminal access again in winbox and some ( not all ) of the static routes have reappeared.
try again 
Don
The lockup Problem appears to occurring only on the PPP client end of the link.
The PPP link is still functioning well but the user interface via winbox is locked out. Tried restarting winbox a few times and got the earlier config of dialog boxes but no data. Eventually ( 2 minutes ) the boxes updated and the earlier routs were still present.
The throughput of the PPP link was still OK thru all this when tested but was not under continuous load so wouldn’t have caused the GUI slowdown.
Cheers
Don
Investigated this a bit more…
Only the PPP client end displays the winbox slowdown / fail problem. Try opening a terminal window & it may show the signon after a while & usually Winbox will disconnect shortly after. Rebooting often helps but not consistently.
I noticed earlier that the actual PPP link was still functioning OK.
I tried telneting to the unit instead and the terminal is responding quite OK by this path.
Watching via telnet I opened winbox and tried to get a new terminal. Usual slo/no response.
The telnet session showed a CPU blip but averaged 2% & 4.6MB free.
Cheers
Don
use only winbox to connect to 133C as this uses a lot less memory as console.
for me it seems that your decribed problem is lack of memory - after some while board runs out of it and then close some program that uses a lot of resources. as result - unexpected crashes and weird behavior appear.
try running BT through the routers not from one board to another.
I have found that if you just leave it for a few hours it will reach a point where it is reachable.
pull all mini-pci cards first.
Once you are able to access it do a system/reset-configuration
That may take some time also.
Once Accessable, again remove any and all packages not required.
As for NTP, it is trying to syncronize, even though the OS has not completed its configuration.
I reset-configuration before upgrading any ROS3Beta.
And Never have a problem.