Coding is hard.One of the reasons v7 takes so long, is that routing programs are being completely re-made from scratch. There is nothing yet to backport.
I think you're being a little...harsh here. The reason why I say that is because generally the people on these boards are NOT the people whom make the decisions at Mikrotik. So giving the support people grief will only generally cause for nothing to get done. It might even be a detriment honestly.Coding is hard.One of the reasons v7 takes so long, is that routing programs are being completely re-made from scratch. There is nothing yet to backport.
Project management is hard.
But this is your job.
Excusatio non petita, accusatio manifesta
Second time, please, stop making fun of us.
Maybe it is just me, but I feel It is disrespectful coming from an employee to a customer.
This tends to backfire, in spectacular ways.
I expect v7 immediately after Christ Second Comingv7 ?
I think it will be a Christmas gift
But what year?I think it will be a Christmas gift
Century 22But what year?I think it will be a Christmas gift
Thanks for the info. V6 is there over 3 yearsRouterOS v6: may 2013 — present (based on Linux kernel 3.3.5)
RouterOS v5: March 2011 — September 2013 (based on Linux kernel 2.6.35)
RouterOS v4: October 2009 — March 2011 (based on Linux kernel 2.6.26)
RouterOS v3: January 2008 — October 2009 (based on Linux kernel 2.4.31)
That is why it is not a good idea to rewrite all major parts from scratch for a new revision.Rewriting mayor parts from scratch while still maintaining stable releases takes time
new routing requires new kernel. a new optional package with a new kernel (and all the other 'optional packages' compatible with that kernel) is just a new version of RouterOSTo rewrite the routing, a sensible approach would be to have new routing code in a new optional package (as was done with wireless)
and keep the old package for the normal user until the new package is stable.
What I mean is they should have released a new version with a new kernel without combining that effort with a rewrite of all the routing, and then do that rewrite independently in an optional package just like the wireless.new routing requires new kernel. a new optional package with a new kernel (and all the other 'optional packages' compatible with that kernel) is just a new version of RouterOSTo rewrite the routing, a sensible approach would be to have new routing code in a new optional package (as was done with wireless)
and keep the old package for the normal user until the new package is stable.
That is exactly what I had in mind.I would rather see another approach:
The release of a 7 beta with the new kernel (let's say 4.4 - the latest LTS) together with the exiting current packages, which should be not such a big fuss, since kernel networking is generally backwards compatible at API level. This would actually test kernel stability in various scenarios.
Afterwards, a progressive rollout of new features is possible, making use of the new features (which are really interesting - e.g. VRF, multipath routing, per flow routing, KVM).
Yes, I think it is much better. An all-rewritten RouterOS 7 will have so many little bugs scattered all over the place that nobodyReleased featureless in parallel with V6, with features added over time allowing the community to test it.
In many use cases it will be good because not every router make use of ALL the features at the same time.
maybe, something like http://www.mellanox.com/repository/solutions/tile-scm/ ?as we won't get ROSv7 anytime soon, anyone with knowledge how to crosscompile for the CCR platform and build a system to run quagga or similar?
OVPN tunnel has no common code with open source implementation of the said tunnel type.Or maybe it's a complete rewrite of OpenVPN? (licence issues?)
This is a great freaking idea actually.Dear support and developers, could you please spend 10-15 min a month to post progress summaries for V7?
I'm not asking for huge detailed report, just very quick summary so we can see some progress going.
+ 1Dear support and developers, could you please spend 10-15 min a month to post progress summaries for V7?
I'm not asking for huge detailed report, just very quick summary so we can see some progress going.
http://forum.mikrotik.com/viewtopic.php?t=70274#p490716Not yet, but v7beta is coming later this year
Minor kernel upgrade? I heard RouterOS 7 is based on FreeBSD, whereas all previous versions were based on Linux?Easy solution - do not make any expectations
We have already posted from time to time, that the biggest change is under the hood (minor kernel upgrade). There is no new GUI or anything. We are also working on a new routing engine.
Actually we are making really cool stuff even in v6. Look at the recent RC versions - interface lists, domain lists for firewall, repeater mode etc.
+1Dear support and developers, could you please spend 10-15 min a month to post progress summaries for V7?
I'm not asking for huge detailed report, just very quick summary so we can see some progress going.
Exactly 2 ( two) years ago your info was that V7 is in alpha stage. Any progress? At least for a beta release....We will release a beta, when it will exist. Currently v7 is in alpha stage, many functions are not completed and non functional. Beta needs at least all functions to be somewhat operational.
Where did they write this ?A few weeks ago MikroTik wrote something about VRF/Linux Network Namespaces. They mentioned that network namespaces "may" be used. VRF is something very fundamental.
But does it support multi thread? Last time I read quaggas docs it says that quagga is planned to support but so does not have the right libs for that...as we won't get ROSv7 anytime soon, anyone with knowledge how to crosscompile for the CCR platform and build a system to run quagga or similar?
True, but at the moment on BGP and IPv6 I'd settle for anything that works without having to set static ipv6 routes to other BGP speakers loopback IPv6 addresses...But does it support multi thread? Last time I read quaggas docs it says that quagga is planned to support but so does not have the right libs for that...as we won't get ROSv7 anytime soon, anyone with knowledge how to crosscompile for the CCR platform and build a system to run quagga or similar?
+1 - that would be very nice, just to have the feeling that you are keeping the users up to date. And to know whether we are more like at the 40% or the 90% progress mark.Dear support and developers, could you please spend 10-15 min a month to post progress summaries for V7?
I'm not asking for huge detailed report, just very quick summary so we can see some progress going.
+ 8.8 million for the MC7304.I too would love to have this to start testing LTE on MC7304.
Was this presented on a MUM? Was it by MikroTik?Any time now guys...
Too many bud lights that day....
Do not mess up IPv6 NAT (wich is must have)[1] and IPV6 masquerade (not needed).You need to talk to your provider. This is insane! We certainly need to avoid getting NAT everywhere for reasonsI have small LAN and my provider gives me IPv6 /64 network, but I'm not able to use it in my internal infrastructure yet (I have few routed LANs with UTP and WiFi).
Also I have bigger LAN, where I do NAT for IPv4 and when something change, I'm not able to fix IPv6 at the same time as IPv4, as I have no NAT for IPv6. This drives me crazy and is a showstopper why to not deploy IPv6 at every place it could be (until RouterOS with IPv6 NAT). Also this is a reason why to not use/suggest Mikrotik at all.
like that. I get a /48 from my provider on a consumer connection, which can be argued to be a bit large, but any
provider should at least give a /60 or /56 to a customer that has a (home) router.
/64 can only be reasonably given to e.g. a rack in a colocation environment where the provider does the routing.
Already in v6+ 8.8 million for the IGMP snooping
I agree that it is useful to have IPv6 prefix translation (1:1 NAT of entire /64 networks) but that is not related to your "my provider gives me IPv6 /64 network"Do not mess up IPv6 NAT (wich is must have)[1] and IPV6 masquerade (not needed).You need to talk to your provider. This is insane!I have small LAN and my provider gives me IPv6 /64 network, but I'm not able to use it in my internal infrastructure yet (I have few routed LANs with UTP and WiFi).
thank you so much, how to configure this feature to work correctly?
Already in v6
*)bridge - implemented software based "igmp-snooping"
In the pdf you referenced from 2014, it talks about, v6.8 I haven't seen anything beyond 6.41 in RC or beta yet in 2017. Where are these being released at?
6.8 was released in Jan of 2014: What's new in 6.8 (2014-Jan-29 15:52)In the pdf you referenced from 2014, it talks about, v6.8 I haven't seen anything beyond 6.41 in RC or beta yet in 2017. Where are these being released at?
.I'm currently switching to another vendor. Cumulus on ONIE + router.Stop asking and switch to another vendor
What vendor has routers with ONIE?I'm currently switching to another vendor. Cumulus on ONIE + router.Stop asking and switch to another vendor
More than 240Mpps of performance, tested.
7Tb/s with prefixes that are loaded on TCAM module.
Regards,
That is a lot of route filters for such a small number of peers !We bought a year ago a CCR1072,
We are using with 4 peers providing us full routing and with more than 800 filters.
After a reboot, it takes more than 2 hours to apply all routes and filters and it's only using 2% of CPU.
Resume: Sh it product.
One peer is IX point, with a lot of members.That is a lot of route filters for such a small number of peers !
The problem is that it has a 72-core processor but the BGP update task is single-threaded so it uses only 1 of the 72 cores.We bought a year ago a CCR1072,
We are using with 4 peers providing us full routing and with more than 800 filters.
After a reboot, it takes more than 2 hours to apply all routes and filters and it's only using 2% of CPU.
Resume: Sh it product.
MAC-based authentication/accounting is broken in general because it is trivial to spoof MAC addresses. The solution is to utilize some higher level of AAA such as PPPoE or WPA2-EnterpriseYou must find a solution to the problem of Internet theft through the Mac
how to configure it ?Already in v6+ 8.8 million for the IGMP snooping
*)bridge - implemented software based "igmp-snooping"
Currently available only in 6.41rchow to configure it ?
[admin@RouterOS-Testing] /interface bridge set 0 igmp-snooping=
no yes
of which year?Dear Santa Claus,
I would like RouterOS v7 for Christmas
2017...I'm hoping for a Christmas Miracleof which year?Dear Santa Claus,
I would like RouterOS v7 for Christmas
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-10.00 sec 22.8 MBytes 19.1 Mbits/sec 41 sender
[ 4] 0.00-10.00 sec 22.0 MBytes 18.4 Mbits/sec receiver
[ 6] 0.00-10.00 sec 14.1 MBytes 11.8 Mbits/sec 38 sender
[ 6] 0.00-10.00 sec 13.4 MBytes 11.2 Mbits/sec receiver
[SUM] 0.00-10.00 sec 36.9 MBytes 31.0 Mbits/sec 79 sender
[SUM] 0.00-10.00 sec 35.4 MBytes 29.7 Mbits/sec receiver
iperf Done.
Not before v7. Even after v7 I am not so sure they will add HW AES support. The CPU does support it, but MikroTik don't seem to want to invest in RB3011.Could we expect hardware acceleration AES for RB3011UiAS-RM?
IPv6 NAT - I bet we have to wait longer than the IGMP snooping people waited before it comes out. . .Could we expect a setup assistent for IPV6 Nat in Router OS?
Not noticed this BGP generated jitter/latency so far. Searched the forum and did not find something regarding this. Did you post/discussed this problem?I think we can safely assume that v7 is some way off, with Mikrotik staying tight lipped on a release date.
Indicating that they themselves have no idea of when it is going to be released.
We are getting to the stage where we are outgrowing our Cloud Core routers, yet we’re only doing 10-100Mb of traffic typically throughout the day! What is killing us is BGP updates, causing jitter / latency spikes on our network… We have a total of 3 full BGP feeds on a total of 4 Mikrotik Cloud Core routers.
Unfortunately, we have lost patience with Mikrotik and we are in the process of exploring alternatives. We won’t be purchasing further Mikrotik products which require BGP.
I really do think they should spell out the limitations of BGP… As on paper everything looks amazing with the Cloud Core routers, however the reality is VERY different.
+1I think we can safely assume that v7 is some way off, with Mikrotik staying tight lipped on a release date.
Indicating that they themselves have no idea of when it is going to be released.
We are getting to the stage where we are outgrowing our Cloud Core routers, yet we’re only doing 10-100Mb of traffic typically throughout the day! What is killing us is BGP updates, causing jitter / latency spikes on our network… We have a total of 3 full BGP feeds on a total of 4 Mikrotik Cloud Core routers.
Unfortunately, we have lost patience with Mikrotik and we are in the process of exploring alternatives. We won’t be purchasing further Mikrotik products which require BGP.
I really do think they should spell out the limitations of BGP… As on paper everything looks amazing with the Cloud Core routers, however the reality is VERY different.
We have the same problem with wirelessI am also looking for an alternative for Mikrotik Wireless, because of missing
- spectral scan
- Nv2 bad performance in 802.11ac
- missing innovation
- missing Sync
After 14 years with MIKROTIK this is not easy. Our hole Netwotk is Mikrotik based,
But they don't here and don't care the needs in Wireless Market, so we have to look to another Brands!
Pity, but reality
And don´t expect fast Release of ROS7,Don't expect anything kernel related to be fixed in v6.
RHEL 7.4 uses kernel 3.10.0-693. How many years are they behind development?we are now at 4.14.xx how long we will see Kernel 4?
Linux Kernel 5.0 Will be Coming in the Summer of 2018
Mikrotik is 4-5years behind development!
right....RHEL 7.4 uses kernel 3.10.0-693. How many years are they behind development?we are now at 4.14.xx how long we will see Kernel 4?
Linux Kernel 5.0 Will be Coming in the Summer of 2018
Mikrotik is 4-5years behind development!
I only have 1 of each, but I'd love to be able to use LTE too.Hello. I have 11 Mikrotik 912 + Sierra MC7304. I too would love to have this to start testing LTE on MC7304. Thank you.
No, I'm still waiting for VirtIO support in v7The most CCR for BGP would sorted out before they will release v7
so the problems also resolve themselves, no more need for ROS v7
So it's time for ROS v7!DNF was released in 2011, so ...
Well, there is always RouterOS v8...And by now it is much too late to go back to the more gradual migration as described above. At least under the designation RouterOS v7.
I think it refers to the option of skipping v7 altogether and then proceed the way described above.Release of v8 will be in ~ 40 years if they proceed with the same speed.
Yes, that's it. I just think V8 should come with a new kernel - and the rest should be improved sequentially. The worst part of a new kernel are the hardware drivers. The rest is, for the most part, high level enough to not be directly affected by it.I think it refers to the option of skipping v7 altogether and then proceed the way described above.Release of v8 will be in ~ 40 years if they proceed with the same speed.
(i.e. make v8 a v6 with current kernel and then incrementally build from there just as done in v6.xx releases)
And yet, they are still able to deliver a working stateful DHCPv6 server, irony much?RHEL 7.4 uses kernel 3.10.0-693. How many years are they behind development?we are now at 4.14.xx how long we will see Kernel 4?
Linux Kernel 5.0 Will be Coming in the Summer of 2018
Mikrotik is 4-5years behind development!
It's pure speculation, but I think that "just switching kernel" might be the problematic part. If they modified too much in the old one (and it looks like a lot of what RouterOS does is not just in userspace), porting everything forward can be hard. Linux is evolving, and it's being worked on by many people, much more than MikroTik has (I guess). So they can either break from it, go own way, and lose all those nice improvements already done by someone else. Or they can try to keep up with changes, port their stuff, but that can be an awful lot of work.When v7 would just have been a move to a new kernel ...
I'm not sure about that. The drivers are kernel bound, of course. So are the NV protocols - with all that TDMA and whatnot. But (almost) everything else? Is just userspace. Even routing: you change it in linux with a userspace command. The kernel may change as much as they want - the interface is still the same.It's pure speculation, but I think that "just switching kernel" might be the problematic part. If they modified too much in the old one (and it looks like a lot of what RouterOS does is not just in userspace), porting everything forward can be hard. Linux is evolving, and it's being worked on by many people, much more than MikroTik has (I guess). So they can either break from it, go own way, and lose all those nice improvements already done by someone else. Or they can try to keep up with changes, port their stuff, but that can be an awful lot of work.When v7 would just have been a move to a new kernel ...
I don't know. Just passing time, with idle speculation.I'm not sure either (and I hope I made it clear enough).
My basic argument is that used kernel was said to be limiting factor for adding some new features years ago. So it would make sense to upgrade it, and not wait for too long before doing so. Not only can the new version already have some wanted features built-in, but also hardware manufacturers are more likely to support current kernel in their (not yet built-in) drivers, if we're talking about newish hardware. If there was only relatively small amount of modifications made by MikroTik, then why hold off for so long?
I'd also say that there's a difference between controlled from userspace and being in userspace, but I can't really go far with that, because that would be too thin ice for me.
I doubt that MikroTik will share exact details, so speculations is the only thing we have. Other than just waiting, but it's a little boring already.
We are also now in the process of throwing out everything Mikrotik running BGP as it cannot handle it.+1I think we can safely assume that v7 is some way off, with Mikrotik staying tight lipped on a release date.
Indicating that they themselves have no idea of when it is going to be released.
We are getting to the stage where we are outgrowing our Cloud Core routers, yet we’re only doing 10-100Mb of traffic typically throughout the day! What is killing us is BGP updates, causing jitter / latency spikes on our network… We have a total of 3 full BGP feeds on a total of 4 Mikrotik Cloud Core routers.
Unfortunately, we have lost patience with Mikrotik and we are in the process of exploring alternatives. We won’t be purchasing further Mikrotik products which require BGP.
I really do think they should spell out the limitations of BGP… As on paper everything looks amazing with the Cloud Core routers, however the reality is VERY different.
to say more: we already moving out from mtik wireless, now after some accidents bgp routers are on the way.
Didn't you hear? Almost all v7 features have been backported to v6!Looking away from RouterOS v6.x running IPv6 BGP [tables]?
Not being able to verify IPv6 routing table (when not main table!) is really a big deal. (But seeing the routes exists by the number the route counter shows. :sigh:)
And doing policy routing rules for anything else than main table in IPv6.
It feels a bit poorly planned/immature(still?) when compared to I can do the WHOLE vrf-mpls infrastructure with IPv4. But it is only partially implemented with IPv6 (vrf's?) in RouterOS v6.x. :/
Still looking for equalized feature set between IPv4 and IPv6 in RouterOS version 7 or 8(?) in 2020 or 2025(?)
@Mikrotik: Would be nice with a roadmap somewhere telling all us customers which year to expect the release of the first release candidate for the next major release of RouterOS. (Currently it feels sort of I'm/we are left out to dry just looking/scraping for news about the next major below the earths horizon and outer space past the planet of Uranus.)
Would REALLY appreciate a lot more openness compared to fx when GitLab.com encountered extended downtime for days due to an engineers non-intended blunder while doing maintenance that fateful day last year (2017).
No. I have not (before now) stumbled across that information.Didn't you hear? Almost all v7 features have been backported to v6!
So looking forward to when RouterOS will support SR !At least we now have kids control and detect internet! Super essential features for routers running BGP in the public internet
I'm sure that this will never happen... Much like IS-IS..No. I have not (before now) stumbled across that information.Didn't you hear? Almost all v7 features have been backported to v6!
So looking forward to when RouterOS will support SR !At least we now have kids control and detect internet! Super essential features for routers running BGP in the public internet
As mentioned several times on this forum, this does not apply as MikroTik did not use the linux community TILE code in any way. They do not rely on the TILE support of Linux at all.If we draw some conclusions:
- Almost everything from 7.x has been backported to 6.x
- MikroTik has been in trouble about GPL
- It's been in development for a long period of time
- Linux has dropped TILE support
Could it possibly be that they're moving to something that isn't GPL licensed?
Just saying, there are options that aren't Linux.
The problem is the old kernel ROS 6 based on, there much improvements only possible with newer kernel.Well if they say they backported most (all ?) v7 functionalities into v6, I don't really see a problem here.
It becomes more clear that the kernel has been so heavily modified that it is not so easy to upgrade it anymore.All new products with IPQ have driver throughput problems, Mikrotik works with a back ported driver, it would be easier to use the origin one,
but it is only available for newer kernels...
here the delay takes its toll
Maybe Mikrotik wait for kernel 5.xLinus Towards launch Kernel 5.xx, happens probably this summer....
right, and then hang on old drivers, that is exactly what happendIt becomes more clear that the kernel has been so heavily modified that it is not so easy to upgrade it anymore.All new products with IPQ have driver throughput problems, Mikrotik works with a back ported driver, it would be easier to use the origin one,
but it is only available for newer kernels...
here the delay takes its toll
This is of course always a tough decision: do you limit the kernel changes and try to forward them to the kernel development group, which will cause a lot of wasted time due to discussion about how to implement some feature, or do you just "fork" the kernel based on some version uptodate at that time, causing the problem there is now.
On the other hand, when you do all the work on a forked kernel, you are not affected by decisions like the dropping of TILE support in the kernel.
so 15 years from version to version. Nice... so we only have another 11 years.DNF was released in 2011, so ...
That is not the problem.I am still not sure why Mikrotik didn't go with IP Infusion....
To be honest, I think it's both. However I think what you bring up is likely more of the issue though.
That is not the problem.
The problem is getting a modern kernel that boots on all of the platforms while supporting hardware offloads.
Just my opinion...
Then explain the difficulty's with migrating on to newer Kernel.....RouterOS uses very little of open source code. MikroTik is in direct relationship with the CPU manufacturers. We use the official drivers and code. We don't use open source code.
If we were using all open source code, it would be easy to upgrade. Now we must only rely on ourselves to upgrade all programs.Do you have any timeline? We are waiting more than 4 year for the VRF supportThis is the difficulty
Is RouterOS 7 going to ship a Linux kernel?This is the difficulty
That is why it is often impractical to develop software that way.If V7 is a major overhaul, I can't image the hour/man needed to do that.
I agree that would be the best way to handle this.Probably a chance to survive this mess is to make a release that aligns the kernel with a reasonably current version and
then work from there on replacement of the subsystems that were announced to be rewritten. However it will be impossible
to release such a change as "v7.0" because there will be a big flood of "you announced it will be fixed in v7 and now we have
v7 and it is still not fixed!" complaints.
Why is Mikrotik so against using open source software? We would have working 802.11ac Wave2, 5 GHz spectral scan, OpenVPN UDP support, more secure smbd, httpd, etc if Mikrotik would stop re-inventing everything in-house and use well-tested open source software. You could combine open source kernel with proprietary modules for hardware offloading etc in much the same way NVIDIA provides proprietary Linux graphics drivers for their chips.This is the difficulty
I don't see all this bloat. Well, not at where we talking: low level, kernel, drivers and servers. Take a look at the RaspberryPi: Full OpenSource, ARM, Linux kernel. This is fdisk usage from one I have access to (default install, with graphic desktop and full updated).One reason is probably that when you use opensource software and keep tracking all the updates, you end up with more and more bloated software that does not fit into a space-limited router anymore.
It works fine on the PC platform where space and other resource usage (CPU) has grown with the code, but on embedded hardware it is a bit more difficult.
Remember the current routers mostly have only 16MB of flash space!
(the previous generation usually had 128MB which was a bit more generous, but to store generic opensouce software for all functions you would need more like 512MB-1GB in my experience)
I think you somehow are confusing MB and GB here. Quite common these days!True, it doesn't have OpenVPN and everything else installed - but it would not use more than 1 GiB. Even if it used 3 more GiB: It would be still using less than ROS.
many people think that if 1 megabyte = 1024 kilobytes, then 1 megabit per second = 1024 kilobits per secondmany people also confuse bit (b) and byte (B) for traffic analysis.
LOL True enough. Things move so fast, we lost track of these things. Nowadays we carry in the pocket a mobile with 4GiB of RAM - and my first computer had 16KiB of RAM... Sometimes is enough to make lost track of things...I think you somehow are confusing MB and GB here. Quite common these days!True, it doesn't have OpenVPN and everything else installed - but it would not use more than 1 GiB. Even if it used 3 more GiB: It would be still using less than ROS.
I don't really buy this argument. Linksys, DLINK, etc all use open source software and have no issues selling $50 routers with enough disk space to support all these packages. With good compilation options and stripping of binaries it's possible to get very small versions of many popular packages. And as you say, Mikrotik used to have plenty of disk space with their products, only recently have they been skimping on the NAND size.One reason is probably that when you use opensource software and keep tracking all the updates, you end up with more and more bloated software that does not fit into a space-limited router anymore.
It works fine on the PC platform where space and other resource usage (CPU) has grown with the code, but on embedded hardware it is a bit more difficult.
Remember the current routers mostly have only 16MB of flash space!
(the previous generation usually had 128MB which was a bit more generous, but to store generic opensouce software for all functions you would need more like 512MB-1GB in my experience)
Well, sometimes those changes are simply not accepted into kernelIf any manufacturer is reading this: STOP DOING OUT-OF-LINE KERNEL RELEASES. UPSTREAM YOUR CHANGES.
What it really boils down to is management decision to go where the money is. This is not unexpected or really all that surprising. Apple does the same thing. What's troubling is that they are effectively alienating some of their fiercest advocates, service providers, in the same way Apple is alienating their professional customers. From where I stand, it's a short-sighted and frustrating turn of events; especially since they aren't really communicating this change in strategy to their customers (maybe in shareholder meetings, but I'm not privy to those).I agree with almost everything @brotherdust has said above, his comments match my own feelings on the current state of Mikrotik.
I have been a huge fan of Mikrotik's RouterOS, but I am using it less and less. The service provider market has kept evolving but RouterOS has not evolved with it.
Mikrotik had a lot of momentum with RouterOS and Service Provider features, e.g. they moved from Quagga to their own Routing engine and added MPLS features during the 3.x and 4.x phase, and then it all just stopped during the 5.x phase. That was over 7 years ago now!
Mikrotik's sales are growing and they are focusing on the areas where they see growth, but that does not mean long term success.
This reminds me somewhat of Toyota thinking they could keep growing by making variations of the Corolla and Camry, so they stopped producing cars like the Celica and Supra, but what they did not realise was that while sales volumes for the Celica and Supra were low, they were the "hero" cars that were attracting buyers to the brand... The result, Toyota, from #1 to #6...
The Service Provider features like BGP, OSPF, MPLS, BNG (PPP/DHCP) are the "Hero" features of RouterOS...
That's a fair point. It's not a simple matter to get those changes accepted. But that begs the question: if they're being denied, why? Maybe write better code? =)Well, sometimes those changes are simply not accepted into kernelIf any manufacturer is reading this: STOP DOING OUT-OF-LINE KERNEL RELEASES. UPSTREAM YOUR CHANGES.
waiting for three year laWe will release a beta, when it will exist. Currently v7 is in alpha stage, many functions are not completed and non functional. Beta needs at least all functions to be somewhat operational.
This makes me seriously thinking about goin back to this U named company, the lack of news on the great version 7 with support for 'old' thing like wave2 support and serious ipv6 support for instance. The Mikrotik hadware is capable but the software is limping on this hardware. Shame on Mikrotik for not even ging us an indication on how far development is. I'm done waiting.Hmm.. heared that U named company released linux 4 based firmware for their routers. Just to mention.
And, as i heared u can have frr on it.
No hands on that equipment, but i keep my eye on it.
And, as i track linux kernel changes, some descent work has been done in networking stack by guys from 6wind and cumulus.
So, mikrotik team, you have a chance. Or had?
The problem is that the U named company has so few options for routers, and so little router functionality in WiFi equipment.This makes me seriously thinking about goin back to this U named company, the lack of news on the great version 7 with support for 'old' thing like wave2 support and serious ipv6 support for instance. The Mikrotik hadware is capable but the software is limping on this hardware. Shame on Mikrotik for not even ging us an indication on how far development is. I'm done waiting.
The real issue is lack of clarity. No communication or anything to keep us posted on the progress of the new kernel version. ( they call it version 7 but that's just a number ).The problem is that the U named company has so few options for routers, and so little router functionality in WiFi equipment.This makes me seriously thinking about goin back to this U named company, the lack of news on the great version 7 with support for 'old' thing like wave2 support and serious ipv6 support for instance. The Mikrotik hadware is capable but the software is limping on this hardware. Shame on Mikrotik for not even ging us an indication on how far development is. I'm done waiting.
the customers answer of missing features, don't buy itfrom what I'm seeing from a business perspective there is no point pushing for the mythical v7 as long as the v6 still sells well
Mikrotik doesn't see v7 as a priority because v6 still sells.
when v6 will become a deterring factor we will see v7 being worked on seriously until that day comes don't expect to see any major investment in v7
you should all that this is the way things are v6 is available as is because there is no official word about v7 or future roadmap take that to consideration when buying Mikrotik products.
v7 the eternal alpha will be a unicorn until maintaining v6 will cost more than developing a new one
true endeedAfter these years, I do not expect any serious answers to come from the Mikrotik. .... Ubiquity is much more flexible in communication ... another region, another morality
I've been forced into ubiquiti for Wireless gear (PtMP Sector -> CPE networks), comfortably finding 27dbi, 29dbi dishes , 20+dbi sectors and a functioning proprietary high-speed AC protocol beat out Mikrotik systems - but FUCK ME if Ubiquity's management tools are absolute shit compared to Mikrotik / Winbox - talking purely over 500+ms links here my ubiquity web-portals are constantly timing out, where I can manage my RB1100's spread out over the same country fine (still use Mikrotik for routing and PPP termination).After these years, I do not expect any serious answers to come from the Mikrotik. .... Ubiquity is much more flexible in communication ... another region, another morality
Router OS v7 is never going to happen.2019?
Well ... so let's get routeros v8.0.0 ...Router OS v7 is never going to happen.2019?
Are you also happy with VRF implementation?Numbering doesn't matter, I am happy with them as it is.
"MikroTik. Destroying the company since 1999. For 20 years already." (c)3) Continue the nose dive, destroy the company
1995 Arnis and John start to provide Internet services to local businesses.
Wireless network project in Moldova.
1996 SIA “Mikrotīkls” is incorporated. The name means “Small network” in the Latvian language.
1997 Starting work on a routing utility, which then is a MS-DOS program and is called “Tik” software system.
PC routers based on 2.4GHz Aironet platforms IC2200.
1999 Installations in Kosovo and the first Linux based RouterOS v2.0Yes, then the Mikrotik hardware would finally openWRT friendly.If MikroTik is unwilling to be transparent at the very least they could do is help RouterOS become Open Source so those who care about the platform can assist in development. Many of us have the skills needed and are willing to sign agreements with vendors to help push Mikrotik ahead.
Can you please define the definition of transparency? I'll save you some time, "the quality or state of being transparent".What do you mean? We have been very open about it.
It is a good topic for jokes but bad choice for serious discussion. Yeah. They flexibly told me that their products will misbehave if power is lost unexpectedly and the only solution is factory reset and restore from backup. Unfortunately I am stuck with them on three hospitality sites...Ubiquity is much more flexible in communication
This is apart of the development process, many of us are the people who build complete platforms and architectures from the ground up. Even if a timeline or development status stated there was work on a specific feature which was later removed during a release then people would be enthralled there's at least some level of transparency.Those are subjects to change. I don't want to promise anything and later tell you that we decided to remove something, or similar.
All network equipment should have UPS units with AVR, to not have conditioning and backup power supplies attached to radio and networking equipment is asking for trouble.Yeah. They flexibly told me that their products will misbehave if power is lost unexpectedly and the only solution is factory reset and restore from backup. Unfortunately I am stuck with them on three hospitality sites...
Can't you be just happy that there are free updates which not only fix security holes but also improve functionality? And you can get all of that with device prices starting at $50?
Without sounding harsh, I think you would have been better off not telling us about v7 if that's the case.Those are subjects to change. I don't want to promise anything and later tell you that we decided to remove something, or similar.
No??We have not advertised v7 as a version.
Just google, site:mum.mikrotik.com "v7"We have not advertised v7 as a version. The first post in this forum thread is by a user.
MikroTik only mentions v7 if there is some feature, that can't be implemented because of kernel limitations. In that case, it's technically not possible to make something until we have new kernel.
Other than that, have we told you there will be a v7 soon?
You know what would help fix this? Having an open source version of Mikrotik. Some of your users are the engineers who develop operating systems, have worked on various OSes including integrating code from KAME in to the BSD projects.MikroTik only mentions v7 if there is some feature, that can't be implemented because of kernel limitations. In that case, it's technically not possible to make something until we have new kernel.
Fake News!We have not advertised v7 as a version. The first post in this forum thread is by a user.
MikroTik only mentions v7 if there is some feature, that can't be implemented because of kernel limitations. In that case, it's technically not possible to make something until we have new kernel.
Other than that, have we told you there will be a v7 soon?
Correct, that indeed is the thing you keep telling us.Sorry to break the news to some of you, but the official stance has always been - use v6. We are working on v7, and it will be released when it will be ready.
Developing on old Platforms often costs more ManPower then needed, there are so much changes in 4.xx Linux Kernel, why invest the time to back port them?Correct, that indeed is the thing you keep telling us.Sorry to break the news to some of you, but the official stance has always been - use v6. We are working on v7, and it will be released when it will be ready.
The problem we have is that for many years now you refer to this V7 and there even is a subforum called V7Beta, this raises expectations that after many years turn into frustration
In that case, please do not say V7 but instead say: Some version we might release in the (probably distant) futureIt doesn't even mean v7 exists.
In that case, please do not say V7 but instead say: Some version we might release in the (probably distant) future
Why not, holds the same informational value: nothingIn that case, please do not say V7 but instead say: Some version we might release in the (probably distant) future
Really?
Sorry to break the news to some of you, but the official stance has always been - use v6. We are working on v7, and it will be released when it will be ready.
Elon has stated publicly the release of information about the model will come out around March 15.When is the new Tesla Y model coming out? What features will it have?
The current kernel needs quite a few modifications in order to accommodate much needed improvements.A lot of the roadmapped features have been put into 6.x because of the delays 7.x brings. 7.x isnt going to suddenly have a heap of new features, 7.x will likely be the latest 6.x but on new kernel
Before he said that. You didn't know about March 15, right?Elon has stated publicly the release of information about the model will come out around March 15.
We've been patiently waiting 5+ years :/Before he said that. You didn't know about March 15, right?Elon has stated publicly the release of information about the model will come out around March 15.
OK so, please wait for a similar statement.
How about you get the big boys at Mikrotik to the table so we can have an adult discussion?Before he said that. You didn't know about March 15, right?
OK so, please wait for a similar statement.
Wait, isnt this poster the magic unicorn? Even the magic unicorn doesn't know when v7beta will come out and thus we are all being treated equally. By the way buset1974 is obviously a student of software engineering and knows of what he speaks, the rest of you yahoos are smart as whips but on big picture stuff not really. Anyway, the real question is what will happen first, v7Beta released or a fricken stable version of Apex Legends.........Hey guys, stop harassing Mikrotik about V7
I am not a blind fan, but in this case, @normis is right. Anyone, who ever did a continuous development, knows that specifying future dates of release is a suicide. If you don't make it on time, everyone will complain (and it will be legit because promise was not delivered). If you rush it and make mistake/bug, everyone will complain (and it will be legit because even a simple bug can cost business millions).
Can't you be just happy that there are free updates which not only fix security holes but also improve functionality? And you can get all of that with device prices starting at $50?
In addition, demanding things this way is a bit ridiculous: "what mikrotik is doing ... is wrong" , "you need to address". Cmon... are you a shareholder or what?
Yeah. They flexibly told me that their products will misbehave if power is lost unexpectedly and the only solution is factory reset and restore from backup. Unfortunately I am stuck with them on three hospitality sites...Ubiquity is much more flexible in communication
Same here I'm willing to join forces towards a community driven project.How about you get the big boys at Mikrotik to the table so we can have an adult discussion?Before he said that. You didn't know about March 15, right?
OK so, please wait for a similar statement.
You're only showing us how worthless the MikroTik training is since there is "no v7", no road map, nothing to show, no transparency and no engagement for the community who would gladly work on an community distribution (sort of like how people can load SwOS).
At least you can repair a broken tesla. But we can't repair RouterOS. BGP has been broken for 5+ years now and nobody knows when it will be fixed.When is the new Tesla Y model coming out? What features will it have?
I'd be interested as well. A community driven firmware could push MikroTik back in to the limelight, but as right now they're falling to the way side.Same here I'm willing to join forces towards a community driven project.
Good idea. Write us the result...We should email the executives of the company starting with Mr. Tully.
Normis:We have not advertised v7 as a version. The first post in this forum thread is by a user.
All unfixed Problems with ROS, where we get the answer, „will be fixed in ROS7“ show us how Mikrotik cares about there costumers, we decided last year to not buy Mikrotik Wireless anymore because of ARM NV2 Issue, what looks like it fixed now. (first issues told to support on Xmas 2017. viewtopic.php?f=7&t=128916&p=633371&hil ... v2#p633371 -to now, so it is about 13 month for this fix) and in all this time nothing as information so I think it is not possible to do serious Business with Mikrotik.Sry Mikrotik - your public relations behaviour is a total desaster when it comes to this topic.
This is a user forum. So I don‘t expect u guys to do the right thing nor do I expect that u even know the truth. We all know that here and there especially software projects might end in a desaster. In combination with the announcement years ago there is a dramatic chance, that ur development is lost in space, i.e. will never been finished.
This would be a desaster for the community but it will hit others even more if their business is based on your products.
Answering ‚it will be ready when its ready‘ is not an option anymore after years of ‚coming soon‘ answers. But this is again something I would expect to be dealt with by you board. So there must be anything more serious than ‚alpha status‘, ‚coming soon‘, ‚not ready yet‘. If you are lost in space and don‘t believe you gonna make it shortly - plz tell us. This will be the better option and will let u determine the right action to be taken.
Nothing has changed from that time. Please suggest what answer you wanted to hear.Normis:We have not advertised v7 as a version. The first post in this forum thread is by a user.
Your post from 08Jan 2015 - v7 is currently in internal testing, and will be released as beta when it's stable enough
Your post from 06 Feb 2015 - We will release a beta, when it will exist. Currently v7 is in alpha stage, many functions are not completed and non functional. Beta needs at least all functions to be somewhat operational.
This topic about V7 was created itself?
And now you say V7 is not a version? Are you kidding?
I understand you can not say anything, but do not make fun of us at all
We are not able to do Business decisions with a maybe or "in near future"
Nothing has changed from that time. Please suggest what answer you wanted to hear.
The feature I'm personally interested in is udp support for openvpn which was requested in 2008. It's not even a new feature just an extension of existing one. It was never stated that it is not feasible or possible.It is not never. I suggested to stop worrying about it and use v6.
About the features, which features do you need? Remember that MikroTik does not implement features by demand. It might not be feasible or possible to make something. Everything has to be evaluated individually.
As a WISP - I'm primarily interested in wireless so I'm interested about spectral scan on AC. And wave2 support?Nothing has changed from that time. Please suggest what answer you wanted to hear.
As many other people, I also expect that when v7 comes out, my router will be able to do anything I've ever dreamed of and more, and make me a coffee on top of that (without any additional hardware required, that's how great will v7 be).... when you now release a v7 that is just a 6.44 with new kernel there will be a large uproar.
The point is that many new features have been postponed for many many years "until v7" even when they do not require a new kernel at all!As many other people, I also expect that when v7 comes out, my router will be able to do anything I've ever dreamed of and more, and make me a coffee on top of that (without any additional hardware required, that's how great will v7 be).... when you now release a v7 that is just a 6.44 with new kernel there will be a large uproar.
Mikrotik writes its own drivers, and maintains its own kernel patches to the Tile CPUs. It is open to discussion if this is a good decision, but I don't know enough (both of the internals of Mikrotik business and of driver development) to have an informed opinion.There is obviously a huge stumbling block as to why a kernel upgrade has not happened, either in a 6.x version or the unicorn v7.
Let’s not forget that Mikrotik is mostly made up of open source software, maybe one or more elements can’t be upgraded to a more recent kernel?
My guess is that it’s fast path / fast track… They both rely on the route cache, which isn’t available in more recent kernels.
some cases promise fix on v7 are not fix yet on v6 thought.Very good response pe1chl, you are getting it.
Anyway, I am just trying to keep expectations mild. Most development is done on v6, as you can see from all the new features, fixes and huge changelogs.
But for now it's RouterOS 9 ("Nein" in German)maybe skip straight to RouterOS X (10)
Don't worry you'll have RouterOS 7 soon enough.. (when pigs will fly)But for now it's RouterOS 9 ("Nein" in German)maybe skip straight to RouterOS X (10)
This is the first question from this thread in 2015.Hi,
When RouterOS v7 beta1 is gonna to be released?
someone may not liveAges are not ending by 2020.
The cycle of RoS?May not... But others will be borned.
)./31 supportIt is not never. I suggested to stop worrying about it and use v6.
About the features, which features do you need? Remember that MikroTik does not implement features by demand. It might not be feasible or possible to make something. Everything has to be evaluated individually.
About features, due to adaptation of IPv6, IPv6 route marking should be implemented ASAP. Also recursive routing should be supported for easy implementation of BGP on IPv6.It is not never. I suggested to stop worrying about it and use v6.
About the features, which features do you need? Remember that MikroTik does not implement features by demand. It might not be feasible or possible to make something. Everything has to be evaluated individually.
my estimation is 1 - 2 yearsGuys, what are you arguing about? Everyone of us knows the v7 release date. Really. Just a mathematics. We are at 6.44 now. Each release requires 3-4 months. So.... 700 - 644 = 56 release cycles. 56 * 4 / 12 = 18,(6) years. So, the year 2037 will be year of 7.00 release!
P.S. Don't thank
You forgot "make me a coffee" button.RouterOS v7 MUST include stable x86-64 architecture for PC. New and old ethernet drivers compatibility, and/or the ability to install custom drivers from linux repositories.
You're right, but lot of people concerns (including me) about the x86-x64 PC support through the forums. So we need a real "official" industrial coffee machine for all.You forgot "make me a coffee" button.
