Ah, I’m not well on sarcasm, try to read twice the post and you understand…
It seems that still the same issue with Dude server package like on 6.46.6 discussed here http://forum.mikrotik.com/t/posts-not-strictly-related-to-v6-49-6-stable/157310/1 Can I update router from 6.49.5 to 6.49.7 without an issue of Dude server OR I have to do it still manulally as described in the post in the link?
Updated CCR1009 and CRS326. Everything is working as it should.
All next version is more consumption CPU.
This is bad road software for me - stop more upgrade. Look this:
https://ibb.co/mb7mTnq
https://ibb.co/YQF6qw0
Really interested what you say regarding new smartphone software.
Hi,
Updated 1 x CCR2004-1G-12S+2XS, BGP router with 14 peers IPv4 and v6. All is fine…
Batch 221 of RBSXTsqG-5acD r2 (SXTsq 5 ac): on some devices there is no wlan1
https://forum.mikrotik.com/viewtopic.php?t=190783
Description
Hello,
We have a CRS 326 24S+ 2Q linked to a Cisco Nexus 9000 using the 40 Gbps interfaces.
SFP are QSFP SR 850 nm.
Everything is working fine until we updated the CRS from 6.48.6 to 6.49.7.
Since this version, FEC options appeared with ROS and link remains DOWN.
Cisco and Mikrotik do not work together anymore. Link is flapping until Cisco decided to bring interface in error-diabled state.
We have tried to turn off FEC instead of auto mode, no link.
None of others options available work.
We tried a 7.6 version of ROS, error remains.
Do you have any idea ?
Thanks for help
Create supout.rif, send it to mikrotik support, not the forum, and repost 6.48.6 while waiting for the fix.
Yes, Thank you rextended
Support is already aware.
Waiting for their answer / fix
Did this change backup, now requiring permission “policy”?
Obviously…
May I ask if there’s a End Of Life date for release 6.x?
Thanks, BR
Officially I don’t think there is one.
My guess is that there won’t be any further v6 release … unless a very severe exploit hits the streets and a v6 release would fix that. And it would have to be a really major one to make MT fix it in v6 since MT is quite aggressive with claims that v7 is mostly on par with v6 in all aspects.
All the functionality developments are strictly in v7 for over a year now and there don’t seem to be any nasty bugs in current v6 releases
Claims, maybe? (where?)
In reality, v7 is not on par with v6 in:
- BGP
- BFD
- routing table queries
- performance on the smallest models
Of course there will be fewer new v6 releases, but they have to remain working and safe to use until v7 is really on par with v6.
Not to be too picky but they have said V7 is actually NOT meant for the smaller models (I am trying to find back some of the posts where it was stated, but can’t find it right away).
While it does work, they have stated clearly this was not the intention. Those models should stay on ROS6.
It will however work for most on the “smaller” models (I have hAP, cAP Lite, mAP, mAP Lite where it works flawlessly for what I need even though those were specifically mentioned as not being capable enough) but when you really stress the config, you will run out of resources.
If you have a device with only 16Mb of memory, forget it. Most likely it will not run.
It’s even mentioned in the specifications for ROS 7
Minimum 32MB of RAM, since RouterOS v7 there is no more maximum RAM.
And that’s nothing new. They have always been quite open about that.
I have never read that from MikroTik. It looks like MikroTik personnel still think there is no problem with the upgrading of hAP lite and mini (32MB RAM, 16MB flash) although we all know the reality is different.
It is also well known that the performance specs in the datasheets for models that were released with v6 do not apply when v7 is installed on them.
I think the performance specs should be labeled with a version (6.xx or 7.xx) or even be split into a 6.xx and 7.xx section for those models that are still being sold.
Now we get angry postings from people claiming “their router gets slower and slower” or “their router refuses to update”…
When MikroTik knew this and would not advise v7, why is a v7 image distributed for those devices (e.g. SMIPS and maybe even MIPSBE)
Don’t have them take away my Hex image, please … (MMIPS) :lol:
SMIPS: the only ones I could find have 32Mb. So that’s on the edge. But that does not mean it can not run. If you keep it simple.
MIPSBE: because a lot of those devices have 64Mb hence they can run ROS7. Within their limits too.
Performance specs not being updated, I understand both sides.
I understand the user side wanting accurate reference info.
But I also understand MT since retesting all those devices with all those test is a GIANT job. Wouldn’t you rather have them improve ROS7 with that time and resources ?
Found the reference.
It’s on the same help page, note 2
https://help.mikrotik.com/docs/display/ROS/Software+Specifications
But again: even though e.g. RBmAPL-2nD is mentioned (mAP Lite), alle devices I have with the same platform, run just fine. With the config I need (once again).
I also found this recommendation:
http://forum.mikrotik.com/t/v7-2rc6-and-v7-2rc7-is-released/156927/1
It can run. But it cannot be updated. MikroTik employees say “it should be no problem” but in practice it is.
MIPSBE: because a lot of those devices have 64Mb hence they can run ROS7. Within their limits too.
Yes, RAM is not the issue here and flash also often is larger on the old models, so upgrading is not a problem either.
But it is slower than before, and users do not understand that.
Performance specs not being updated, I understand both sides.
I understand the user side wanting accurate reference info.
But I also understand MT since retesting all those devices with all those test is a GIANT job. Wouldn’t you rather have them improve ROS7 with that time and resources ?
I think MikroTik has several employees, but not all of them are qualified for the same task.
It seems at this time nobody is qualified to work on the routing code. Likely the employee(s) who coded this is/are no longer available, and development halted. New employees are studying the code and making very small changes a step at a time, but they can not yet take the big job of making v7 on par with v6.
That does not mean there isn’t anyone who can fire up the test rack and performance test equipment and run the tests again. That is quite a different type of task and most likely another person doing it.
Bits and pieces here and there, but mostly claiming v7 can do almost anything v6 does. You know very well how it goes. And yes, many agree that currently v7 is not universal replacement for v6 (in some cases it is, in some far from it). But the attitude by (most?) MT staffers makes me believe they won’t put any more effort in v6. And there’s some reasoning behind that (e.g. why spend time on improving BGP in v6 if BGP in v7 desperately needs attention).