Do you think it would be possible to maybe stop entirely what you folks are doing in terms of advancing ROS with new features etc... I think the most at this point that ROS users wish for in 2012, is that all resources in MT engineering solve the current problems that ROS has been having since the introduction of 5.xx .

Port flapping - This has NEVER been fixed..! Attempts at masking the problem, but MT cant even recreate it, thus they cant fix it. But we are several with the issue since who knows when.

Ethernet ports locking up - Never been fixed or acknowledged as a problem just like the port flapping. Recent RB 411,433,435,493 board are locking up on ethernet ports. In the early 5.xx releases, it seemed to only affect ethernet port 2, but the issue affects all ports and even when downgrading there are still lockups.

Ethernet port mismatch - 5.xx. In my case, I can rarely get ethernet ports to match up. Either one is in full, other auto, or vice versa, they always report the wrong info.. If I do get 2 boards on the same settings, then there is major packet loss.. The only way there is no packet loss is when one is on auto and other with no auto. Combination's like that, but never works when set as suppose to with above mentioned boards.

Its obvious that the major issues that are bringing my company down in stability and customers, is 5.xx. I have a few hundred RB's in service and my network was stable as a rock with first Gen 133's and 433's with 3.xx, precisly 3.30. Went 2 full years without a software related issue. 4.17 was not bad accept that power/signal levels were down compared to 3.30 and some links were not as stable, therefore they got turned back to 3.30. I have some critical RB's that could really use 5.xx tdma, but I couldnt DARE upgrade them or touch them.. Im wayyy 2 scared. Bench testing means nothing as most of these port related issues happen only with some good traffic.

My company has gone from the most stable wireless network most ppl have ever seen and they cant beleive its wireless... Too, the most unreliable wireless network they have ever seen... Customers dropping like flies, ppl getting second connections, VoIP going to other providers and ifcourse, loss in revenue... I just cant understand why MT cant test there software and release it only when its actually stable. Like STABLE!!.

In my short experience with MT, I can only say that there are 2 versions that are actually stable. Well 2 and a half. 2.95, 3.30 and 80% 4.17.

There are issues in these newer RB ports that werent there in the past. Even new boards running 3.30 are having these ethernet port issues.. There has to be something that changed in hardware or code for ethernet ports since the days of 3.xx and 2.xx. Please go back to that hardware and or code.. PLEASE... Im not even ranting here. Im expressing my sincere disappointment in ROS since 5.xx. If I would have know that my network would go from yeahh to neahh in less than a year due to ROS, I would never have built the network based on MT.

MT has great vision, great features, the best in my opinion.. But they in my opinion are going downhill. If a company cant release software that is stable in ALL circomstances, than whats the point.. If you folks cant test your SW enough before releasing it as stable, then dont release.. If you want to continue to release SW and rely on the community to test it and report bugs, then release as beta!..

Tell me ONE release that you have made in the last 2 years that hasnt had a medium to major bug in it.?? Every single release has an issue.. I understand that it cant always be perfect, but why is it other companies can release software with a hell of alot less bugs than MT? Since 5.xx were talking serious issues here, the damn ports dont pass any traffic.! Like thats major.! loll

If I were the CEO of MT. I would be embarrased that we couldnt find the issue. I would stop all other development and focus on resolving this issue and only this issue.! Whats the point in developing further if theres always a hidden major bug that is following each release.! You folks should also quit telling ppl to submit tickets when its issues that have clearly already been submitted.. You folks always say the forum is not for official support so we shouldnt rely on it and submit a ticket with file. Fine, thats how bugs are found, but you folks also should not rely on community members for testing your releases when they are depending on your product for the food on there table. Thats your responsibility to test, test, test and re test., Or atleast release as beta, then ppl like me cant complain.

Im having so many port related issues and my network is so un reliable now, that when im down, my customers dont even call anymore. They just sit there and hope its not too long, or they just leave all together like they have been doing.

DO you folks not feel embarrassed at all??

I personnel could not sleep at night knowing that ppl are having serious issues with your product and you cant even find the problem.. PPl are depending and paying you. You folks just hope the problem goes away, or just that ppl affected just get tired of repeating themselves. You are willing to let those ppl go and hope that the ppl not having issues continue to buy your products. The problem is, that eventually, they will have the same issues also if not fixed.

To my knowledge, there has never been a specific comment on what they think the problem is in regards to port flapping or freezes. Do you folks even have an idea? Or you think there is no problem? I cant understand how a group of extremely intelligent ppl that have developed some great stuff in the past cant figure this one out... These issues will effectively ruin your rep and sales. Some of us have to convince ourselves not to move our entire network to UBNT. I even trying to acquire licenses and 10k equipment to replace MT.. But I love MT.! But since 5.xx, it SUCKS HARD!

Im sure if the entire team actual took our problems seriously, it would be resolved 100% in a matter of days.

Ifcourse I will get deleted again.! WIll I have to resort to signing up again with different usernames and or IPs to get this message across to the roght ppl?.. I hope not. Cause I LOVE MT.. But this cant go on.! IT CANT.

TEST AND FIX YOUR ISSUES REGARDING ETHERNET PORTS. Ifcourse theres now the kernal failure issues... I give up...

AM I the only one this frustrated at MT? This post is only to hopefully motivate the team in fixing thiese important issues and no more.. I dont want war here, I dont want excuses, I dont want to be deleted. I just want MT to work like it use too. You know, when you could impress 20 year engineers with your products and its flexibilty and reliability..

AM I asking too much?

Happy Holidays!

Port flapping - This has NEVER been fixed

There is no such software problem. If you are getting these symptops, check you cables, connectors, interference and ports. You don't see the issues in older versions, because logging of ethernet status was not implemented. This doesn't mean you didn't have the same problem, you simply didn't have any logs about it.

TEST AND FIX YOUR ISSUES REGARDING ETHERNET PORTS. Ifcourse theres now the kernal failure issues... I give up...

Have you contacted support regardin autonegotiation issues? Which ticket number did you get, I can check status.

This is happening on dozens of my boards, as discussed by other users. This is a hardware/software issue. Older boards do not get packet loss, drop in bandwidth between two boards via ethernet, mismatch in ethernet link status and port flapping.. You see, this is the problem. I can find dozens of posts of users discussing these problems, but your answer is always, there are no problems must be something else.!


You have to stop with this.! Yes 3.xx and prior did not report port flapping, but they didnt have packet loss, freezing, port mismatch ports only working when one full and other auto and all the other things I mentioned.. I mentioned this many many months ago before alot of other ppl and you gave me the exact same answer.

Answer this question for me please.. Why since 5.xx all these problems happen, and with older boards if I put back 3.xx, all is well, regardless if port flapping was show in the logs before or not.. No packet loss, no port mismatch and no port freezing.?? Please answer this directly.

One would think after 6 years using your products, I would have a good idea of what I am talking about, and have enough experience to rule out cable intereference problems etc loll..

Answer my question please.. Why before 5.xx no one else complained about these issues..

To fix this, you must first acknowledge that there is a problem.. Just admit it!

You arrogance is really going to start to make smoke come out of my ears.. Where not all stupid yo know.!

Im so sick and tired of answers like you just gave me, that I just called in quotes to light up all my towers with fiber. If I can get atleast half of my network of MT, that would be a twice stable network.. I have to get my back-haul's of your stuff... In the past this was not an issue because tdma wasnt as widely used and there was not as many wireless devices in my area. Now, with cheap ubnt and more ppl learning what can be done at lower costs, sometimes tdma is absolutely necessary now..


Go search your forum here and in any other language and read up on users have issues with the ethernet ports..!!!!!!!

PLEASE. because I love MT

Have you contacted support? I guess your problem could be elsewhere. What is your ticket number? What was the result from contacting support?

Normis about "Ethernet ports locking up" I found that:
1. It happens on RB435G but not on RB600A. Might be related to that switch chip in RB435G???
2. Only happens when ethernet port sends traffic. If it only receives never had happened. Might be ethernet or switch chip Tx counters/buffers overflow ???
3. When it happens disable/enable the ethernet port doesn't resolve it. But toggle the state of "switch-all-ports=yes or no" makes the port work again, without a reboot.
I can send you supout generated when working and when not, but I doubt it will help.

I have wrote about it here
http://forum.mikrotik.com/viewtopic.php?f=1&t=57215

And finaly I think mm690 have the right to complain, so giving him karma.

YES I AGREE..........!

fix the problems that's never been fixed

You arrogance is really going to start to make smoke come out of my ears.. Where not all stupid yo know.!

heh, i remember at MUM Budapest somebody asked about this, and Normis changed the subject.

Be sure, they know about the problem, but they can't find a decent solution to solve it.
May be some cheap chip that does this and exchanging will cost more to manufacture,etc.

They reached a level when they have to invest in brain power more than in advertising power... but the numbers look so good, and the bonuses are so fat that it's a hard decision to make...(fix bugs vs go on)

My "resolution" to Mikrotik:

- Mikrotik, what's your problem ? Do you really think that we need Groove,Omnitik, SXT , etc ?
For most of us, Mikrotik=RouterOS. We need RouterOS to be rock solid, trustfull and be able to sleep well at night, knowing that our routers are ok. I myself stated a few times my problems with hotspot, solution being rebooting the device. Rebooting , of fixing a software problem with a hardware solution...( I even have GSM controlled reboot for RB's)

All these new products, running the same buggy software

For my side, I use MT only for routing, wireless is UBNT or TP-Link (they fullfill my needs)
Most of my wireless devices have uptime in months, but on every RB gateway, i have to reboot them every day.

I know that the decision-making staff is not looking here, but at least you can suggest so shift on company's policy.

I've seen companies going down from the same resons like these presented here. Don't be blind.

heh, i remember at MUM Budapest somebody asked about this, and Normis changed the subject.

This is not true. Please don't invent such things. This issue was not even active on the forum until around may-june (much after the MUM).

My problems i havent reported cors most i sorted myself out, but for real try to sort them out

eg:

queue problems clash with new versions , not stable like in hotspot
userman problems NB* pppoe usage limits dont work perfect like hotspot
hotspot is the only thing that works smooth with queue & bandwidth
making router capable of sending rewrites instead of onboared nand to send to usb for example or networkly to a harddrive ( this concerns life of routerboard )

the rest such as firewall,proxy etc... are all ok..no problems

I have that problem with autoneg on rb1100 with Licensed NEC Pasolink Radios. If We turn autoneg off and set full duplex and full. the Nec port gets Up but RB never get up. We tried even with machine made certified patch cables and no go. And that's happens randomly, like once a week, or 20 times in one day.

Another problem is about stuck routes. If a bgp peer goes down, some routes from the dead peer still works invisible. We cant find those routes, and new ones from the backup peer shows as active in /ip route , but they never get really used.. to fix that issue we have to disable and reenable the dead peer interface or reboot route.

Never heard back from support.

Ticket#2011122366000791

I always said UBNT and MT should have merged. Ubnt has excellent hardware design, MT has routeros.

With ubnt now going public and developing their own EdgeOS based on Vyatta I feel it's too late for a merger, and MT really has to up the ante to compete.