I must also sign under this.Thanks Mikrotik upgraded 2 2011 at 5.25 to 6.7 and 1 completely died and 1 had to be powercycled twice to get it back and I am assuming that will die but am scared to downgrade to 6.6 because it will probably die. I only upgraded because 2011s are so buggy on 5.25 and anything else 5. Should have just upgraded to 6.6 in retrospect. Really getting sick of Mikrotik its just so cheap and Router OS is so awesome its hard to get away. After the 450g disaster of all those capacitors blowing I really question who is running the show over there in Latvia. Talk about trying to ruin a good thing with terrible hardware decisions and quality control.
Please install RouterOS v6.7 make a support output file and send it to support@mikrotik.comI want to say that , maybe the problems what I had , was ROS not support intel i350 network adapter as well .
Just like in the photo, several interfaces with the same MAC address.
I had tried v6.x and v5.x , same problems.
I had tried reset configuration , no any config , but not help.
Sadly I cannot disagree with this.Talk about trying to ruin a good thing with terrible hardware decisions and quality control.
Please install RouterOS v6.7 make a support output file and send it to support@mikrotik.comI want to say that , maybe the problems what I had , was ROS not support intel i350 network adapter as well .
Just like in the photo, several interfaces with the same MAC address.
I had tried v6.x and v5.x , same problems.
I had tried reset configuration , no any config , but not help.
What exact Intel I350 adapter are using using - is it a PCIe card or built-in module on the board?
Resetting the MAC address to the original MAC address will help with your issue.Please install RouterOS v6.7 make a support output file and send it to support@mikrotik.comI want to say that , maybe the problems what I had , was ROS not support intel i350 network adapter as well .
Just like in the photo, several interfaces with the same MAC address.
I had tried v6.x and v5.x , same problems.
I had tried reset configuration , no any config , but not help.
What exact Intel I350 adapter are using using - is it a PCIe card or built-in module on the board?
PCIE card , support output file had already send
This can be done with:Resetting the MAC address to the original MAC address will help with your issue.
/interface ethernet reset-mac-address ether1
v6.7 hasn't been around that long, to make such graphs. I would guess you run an older version then?From version to version is not happening, no changes in the protocol PPP. After 2-3 weeks of continuous operation in the PPP interfaces winbox open very long. Memory grows continuously until you reboot CCR.
With the new 6.7 version. Accordingly dates for December.v6.7 hasn't been around that long, to make such graphs. I would guess you run an older version then?From version to version is not happening, no changes in the protocol PPP. After 2-3 weeks of continuous operation in the PPP interfaces winbox open very long. Memory grows continuously until you reboot CCR.
Resetting the MAC address to the original MAC address will help with your issue.
As you can see in the v6.8rc1 changelog, there is a bugfix for address list issue. Try the pre-release.have anybody noticed the issue with not working address list items on CCR?
we installed CCR-1036 a few days ago, and today I saw two users without Internet access. after disabling/enabling their (static) address list entries, they started working again. I'm surprised because similar problem was fixed more than five years ago in v3.16 (Ticket#2008081866000458), and we don't have this behaviour on our x86 routers
hope so, Normis, but I don't create duplicate address list entries =)As you can see in the v6.8rc1 changelog, there is a bugfix for address list issue. Try the pre-release.
I have seen this and was checking if my actual firewall rule was wrong. Thanks for bringing it up.have anybody noticed the issue with not working address list items on CCR?
we installed CCR-1036 a few days ago, and today I saw two users without Internet access. after disabling/enabling their (static) address list entries, they started working again. I'm surprised because similar problem was fixed more than five years ago in v3.16 (Ticket#2008081866000458), and we don't have this behaviour on our x86 routers
Any of you guys with memory leakage use Radius Manager by any chance? We had the same issue - removed Radius Manager and the API for CPE signal polling, and the memory problem was solved...I have the same problem as OlegM, but much faster
Why, then, need a router if it can not use radius and api. This basic service for authentication and control of users on the Internet. So it is necessary to solve this problem and not exclude it.Any of you guys with memory leakage use Radius Manager by any chance? We had the same issue - removed Radius Manager and the API for CPE signal polling, and the memory problem was solved...I have the same problem as OlegM, but much faster
/TB
You don't really answer the question, but we suspect the API to be the sinner. If the API have session issues - like opening a new session every time, without closing the previous - I assume that could explain the leakage. We could never run for more than 25-30 days on our RB1100, before it ran out of memory and needed reboot. We looked everywhere and tried all possible settings and versions without any improvements. Finally we switched to another Radius without API for signal polling (using SNMP instead) - and problem was gone. Simple test for you to run; just turn off API for a few days...Why, then, need a router if it can not use radius and api. This basic service for authentication and control of users on the Internet. So it is necessary to solve this problem and not exclude it.Any of you guys with memory leakage use Radius Manager by any chance? We had the same issue - removed Radius Manager and the API for CPE signal polling, and the memory problem was solved...I have the same problem as OlegM, but much faster
/TB
FYI - You can explicitly close an API connection by issuing a "/quit" command every time you're done with it, thus not causing leakage.If the API have session issues - like opening a new session every time, without closing the previous - I assume that could explain the leakage.
It turns out it is a hole,in which the missing session are not disabled themselves.FYI - You can explicitly close an API connection by issuing a "/quit" command every time you're done with it, thus not causing leakage.If the API have session issues - like opening a new session every time, without closing the previous - I assume that could explain the leakage.
But if you "/quit" it, it's no longer leaking, right?It turns out it is a hole,in which the missing session are not disabled themselves.FYI - You can explicitly close an API connection by issuing a "/quit" command every time you're done with it, thus not causing leakage.If the API have session issues - like opening a new session every time, without closing the previous - I assume that could explain the leakage.
Turns out, that remains in the memory. For the memory is constantly growing.But if you "/quit" it, it's no longer leaking, right?
(Or do you mean that even quitted connections leak memory?)
POE-out port flapping occurs in 6.7 (maybe others) when any interface examination is being done. This includes the "Interfaces" window being open in Winbox. If you close this window (and any other window that would "examine" ethernet ports) flapping will disappear. If you use ssh, and run POE monitoring, you will see the same flapping. This is fixed in the (unreleased) 6.8.Upgraded today on RB750UP, previous firmware was 5.24, suddenly I have port flapping, but only if I try to connect with winbox, if I use ssh everything is ok. I have some devices powered on ports, try to change poe status to forced-on and auto but nothing change. I upgraded also rb firmware, but still the same issue. I am connecting to in with winbox on linux (wine).
Any other suggestion?
Also having issues with the fast ethernet bank on 2011's with 6.7, no ethernet activity with some devices it will not negoiate a state. Have not had to much time to look into it as we are to busy to worry about it now.10mbit leds are not working on 2011
i get no led activity on 10mbit devices on the fast ethernet side. gig side it works fine?
any update? ticket has been sitting there for weeks now, along with ups issues
make sure that you have dns server specified in the router and you have access to the internet from the router.Could you please show me how to upgrade, it says could not resolve address upgrade.mikrotik.com
We're having the exact same issue, appears it may have been happening for several versions (upgraded from 5.X to 6.7), as we downgraded to 6.6 and the issue is still present (don't want to keep downgrading).All routes with routing mark is down after upgrade from 6.6 to 6.7, remove the routing mark will make route establish
Could you tell me how? I'm new with RouterOS, but when I clicked packages->Check for Upgrade or AutoUpgrade the System told me the newest Version is 5.26 !?? I'm using a new RB-951G-2HND I was able to upgrade from 5.24 to 5.26 but not higher, my License told me Upgradeable to v7.x???yes!Is it possible if i upgrade from ROS V5.26 to ROS v6.7 directly ?
PPPoE users are authorized to RB1100AHx2 through Radius request to billing. Found that RB1100AHx2 not send requests (Accounting-Update) on billing. Billing believes that the user disconnected and hang PPPoE session. In the settings billing parameter "Timeout hung session (s) = 60 sec". If RB1100AHx2 not send (Accounting-Update) on Billing - Billing disconnect PPPoE session number. The problem occurs only on 6.8rc1 (2014-Jan-21 11:25).Version 5.25 (5.26) works perfectly and user not hanging.
Can anyone confirm this error?
when it is donewhen comes 6.8?
Download http://download2.mikrotik.com/routeros/ ... be-6.7.npk file.Could you tell me how? I'm new with RouterOS, but when I clicked packages->Check for Upgrade or AutoUpgrade the System told me the newest Version is 5.26 !?? I'm using a new RB-951G-2HND I was able to upgrade from 5.24 to 5.26 but not higher, my License told me Upgradeable to v7.x???yes!Is it possible if i upgrade from ROS V5.26 to ROS v6.7 directly ?
The same problem.Routeros 6.8rc1 isn't accounting the amount of traffic, on radius always have 0kB in database.
the same questionHave someone tested radius accounting with that version(6.8 jan29)?
Solved in 6.9Yes, very intersting(Radius Accounting(Interim-Update) in PPPoE). Bug is solved or not?
What is port isolation on MT?fixed port isolation on CRSs
do you have DST maybe ?BUG report
Version: 6.9 30/01/2014 13:02:02
plateforme: RB1100AHX2
DateTime in log is not logged with current system time. It seems recording in GMT
Instead GMT+1 in my case...
I don't knowNormis,
Why is everybody talking about 6.9 on this forum.
And not 6.9 beta of 6.9rcx.
Let me fix this for you:We have not released any new version since v6.7 !
Not hours, we did not put it on the website at all. Only torrent was released for some 5-10 minutes, then files were deleted.Let me fix this for you:We have not released any new version since v6.7 !
"We have released v6.8 since v6.7 but pulled it back from the website after a few hours due to RouterBOOT problems"
Normis,do you have DST maybe ?BUG report
Version: 6.9 30/01/2014 13:02:02
plateforme: RB1100AHX2
DateTime in log is not logged with current system time. It seems recording in GMT
Instead GMT+1 in my case...
I have the same issue on RB2011 with version 6.7im have strange bug from 6.5-6.8rc1
on 6.4 all is well
double traffic per minute. using ospf+mpls, no difference track system
on or off mtu is 1518 or 1508, double traffic is only on transport
port....