documentation draft:Nice changelog, I don't know how to find that "rose-storage" ..
I've downloaded the extra packages but somehow I've missed it, now I've found it :) thank you!you need to install rose-storage package first
For a CHR instance without an active licenseWhy do you quote whole preceding post? Does it help answering? Do you repeat what your interlocutor says when you discuss?
Intresting. Mikrotik planning to enter the SAN/NAS business?documentation draft:
https://help.mikrotik.com/docs/display/ROS/ROSE-storage
16:43:05 container,info,debug importing remote image: zabbix/zabbix-proxy-sqlite3, tag: alpine-6.0-latest
16:43:05 system,info item added by cesar
16:43:07 container,info,debug error response getting manifests: 404
16:43:07 container,info,debug was unable to import, container 4a07240c-862b-4861-a16a-68605478ad54
16:45:28 container,info,debug importing remote image: zabbix/zabbix-proxy-sqlite3, tag: alpine-6.0.12
16:45:28 system,info item added by cesar
16:45:31 container,info,debug getting layer sha256:6875df1f535433e5affe18ecfde9acb7950ab5f76887980ff06c5cdd48cf98f4
16:45:32 container,info,debug layer sha256:6875df1f535433e5affe18ecfde9acb7950ab5f76887980ff06c5cdd48cf98f4 downloaded
16:45:32 container,info,debug getting layer sha256:2068be5b412156c5bc2936aeb988446cb6ac458c4c408ac51b5143e9632073f0
16:45:33 container,info,debug layer sha256:2068be5b412156c5bc2936aeb988446cb6ac458c4c408ac51b5143e9632073f0 downloaded
16:45:33 container,info,debug getting layer sha256:35af6ce2b615d78f6617ef90fdbb0aef91a77c766594c28325a8e9e589d0e002
16:45:33 container,info,debug layer sha256:35af6ce2b615d78f6617ef90fdbb0aef91a77c766594c28325a8e9e589d0e002 downloaded
16:45:33 container,info,debug getting layer sha256:7becd6903f60f84a63358dbfbf033e34094e07d255085fe0d9a2fe48481e74b6
16:45:34 container,info,debug layer sha256:7becd6903f60f84a63358dbfbf033e34094e07d255085fe0d9a2fe48481e74b6 downloaded
16:45:34 container,info,debug getting layer sha256:21bb24f368b7ae4b135a1ef432a6379a54310c37e8a7b8d54d0260d7cd768f9d
16:45:35 container,info,debug layer sha256:21bb24f368b7ae4b135a1ef432a6379a54310c37e8a7b8d54d0260d7cd768f9d downloaded
16:45:35 container,info,debug getting layer sha256:9e1e869413aec50921ae70ba3b2098e56ab598bb6a26d2b0d5c697f7c433cb00
16:45:37 container,info,debug layer sha256:9e1e869413aec50921ae70ba3b2098e56ab598bb6a26d2b0d5c697f7c433cb00 downloaded
16:45:38 container,info,debug getting layer sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1
16:45:38 container,info,debug layer sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 downloaded
16:45:38 container,info,debug getting layer sha256:e4034d2118985bc524c23df0d8c998c604ee97aef464494c39111bf32ebd9335
16:45:39 container,info,debug layer sha256:e4034d2118985bc524c23df0d8c998c604ee97aef464494c39111bf32ebd9335 downloaded
16:45:39 container,info,debug import successful, container c3a27c76-186a-47bf-ace4-04fcff0790fd
https://www.youtube.com/watch?v=BbDnBxlBTdYZero Trust Cloudflare package option missing.
First, one shouldnt feed the troll posts like mine ;-PPWhy do you quote whole preceding post? Does it help answering? Do you repeat what your interlocutor says when you discuss?
Yes you're right, 10 NVMe slot is not for a router but for a server. ;-)The RAID examples in the docs (https://help.mikrotik.com/docs/display/ROS/ROSE-storage) use NVMe as examples. One could expect to see more 2x16's with more RAM and NVMe slots--or possibly some 2x32's, with 32 cores, 32GB of RAM, 2-4-more NVMe slots, and 40G or 100G networking.
Yep, couldn´t agree more, I am disappointed by 7.8 as well: people are not waiting for ROS to have perfect storage support (ROSE). Having containers is great, but it should not be the focus. I am waiting for solid WIFI features (band steering, roaming, wireless bridging for wifiwave2, single capsman to manage old and new wifi), better IPv6 support (fast track, rfc9096), more HW acceleration.This is a disappointing release as 7.8, should have been 7.7.1.
When are we going to see Mikrotik address those critical route/switch features that most enterprises use. Specifically:
RouterOS version 7.8beta2 has been released "v7 testing" channel!
*) certificate - improved certificate management, signing and storing processes;
/certificate/add name="r1-ca" common-name="r1-ca" subject-alt-name="email:r1-ca" key-size=prime256v1 key-usage=key-cert-sign,crl-sign
:do {/certificate/sign [find name=r1-ca] name=r1-ca} on-error={:delay 3}
/certificate/add name="r1" common-name="192.168.2.14" subject-alt-name="IP:192.168.2.14" key-size=prime256v1 key-usage=digital-signature,content-commitment,key-encipherment,key-agreement,tls-server
:do {/certificate/sign [find name=r1] ca=r1-ca name=r1} on-error={:delay 3}
/certificate/add name="r1-r2" common-name="r1-r2" subject-alt-name="email:r1-r2" key-size=prime256v1 key-usage=digital-signature,key-encipherment,data-encipherment,key-agreement,tls-client
:do {/certificate/sign [find name=r1-r2] ca=r1-ca name=r1-r2} on-error={:delay 3}
:delay 2
/certificate/export-certificate r1-ca file-name=r1-ca
/certificate/export-certificate r1 file-name=r1
/certificate/export-certificate r1-r2 file-name=r1-r2 type=pkcs12 export-passphrase=passphrase
/ip/pool/add name=r1-r2 ranges=192.168.1.2
/ip/ipsec/mode-config/add address-pool=r1-r2 address-prefix-length=32 name=r1-r2 split-include=0.0.0.0/0 system-dns=no
/ip/ipsec/policy/group/add name=group1
/ip/ipsec/profile/add dh-group=ecp256 enc-algorithm=aes-256 hash-algorithm=sha256 name=profile1 prf-algorithm=sha256 proposal-check=strict
/ip/ipsec/peer/add exchange-mode=ike2 local-address=192.168.2.14 name=peer1 passive=yes profile=profile1
/ip/ipsec/proposal/add auth-algorithms=sha256 enc-algorithms=aes-256-cbc,aes-256-gcm lifetime=8h name=proposal1 pfs-group=ecp256
/ip/ipsec/identity/add auth-method=digital-signature certificate=r1 generate-policy=port-strict match-by=certificate mode-config=r1-r2 peer=peer1 policy-template-group=group1 remote-certificate=r1-r2
/ip/ipsec/policy/add dst-address=192.168.1.0/24 group=group1 proposal=proposal1 src-address=0.0.0.0/0 template=yes
/certificate/import file-name="r1-ca.crt" name="r1-ca" passphrase=""
/certificate/import file-name="r1.crt" name="r1" passphrase=""
/certificate/import file-name="r1-r2.p12" name="r1-r2" passphrase="passphrase"
:delay 2
/ip/ipsec/mode-config/add name=cfg1 responder=no
/ip/ipsec/policy/group/add name=group1
/ip/ipsec/profile/add dh-group=ecp256 enc-algorithm=aes-256 hash-algorithm=sha256 name=profile1 prf-algorithm=sha256 proposal-check=strict
/ip/ipsec/peer/add address=192.168.2.14/32 exchange-mode=ike2 name=peer1 profile=profile1
/ip/ipsec/proposal/add auth-algorithms="" enc-algorithms=aes-256-gcm lifetime=8h name=proposal1 pfs-group=ecp256
/ip/ipsec/identity/add auth-method=digital-signature certificate=r1-r2 generate-policy=port-strict match-by=certificate mode-config=cfg1 my-id=dn peer=peer1 policy-template-group=group1 remote-certificate=r1
/ip/ipsec/policy/add dst-address=0.0.0.0/0 group=group1 proposal=proposal1 src-address=0.0.0.0/0 template=yes
09:02:34 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:283ba582c62ec4fa:57b2c5210d7931a4
09:02:34 ipsec,error can't get private key
09:02:34 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:283ba582c62ec4fa:57b2c5210d7931a4
09:02:36 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:1e4c321c0f62bbb0:b86fa5304054f5df
09:02:36 ipsec,error can't get private key
09:02:36 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:1e4c321c0f62bbb0:b86fa5304054f5df
09:02:37 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:005d4fe0daeadb1d:5319ddb77408abbb
09:02:37 ipsec,error can't get private key
While I agree with your point, of course this is NOT a release. It is not 7.8.This is a disappointing release as 7.8, should have been 7.7.1.
When are we going to see Mikrotik address those critical route/switch features that most enterprises use.
My hAP ac2 (running 6.49.6, but nevertheless) has 696kiB of free space on flash. So nothing new for 7.8beta ...By now my hAP ac2 has only 1452 KiB of free space on the flash
I agree. I think Mikrotik has to decide what segment to focus on. Home users that demand bleeding-edge all-in-one boxes or enterprise that need proven solutions and functions.This is a disappointing release as 7.8, should have been 7.7.1.
When are we going to see Mikrotik address those critical route/switch features that most enterprises use. Specifically:
1. BFD fixed
2. BGP-VPNv4-VRF RR fixed
3. Something equivalent to Cisco DMVPN, HP DVPN, Meraki AutoVPN, or Fortinet ADVPN, etc.
4. EVPN
5. MPLS Fast Reroute
6. BGP Multi-path
7. L3HW off loading that is compatible with MLAG and VRRP
8. L3HW off loading for VXLAN
9. L3HW off loading for QinQ
Until those are delivered I can’t use Mikrotik in any medium or large US companies I support. Cant use it in the data center, can’t use it in the WAN edge, can’t use it in the LAN.
ROSE is nice but it’s not helping me get Mikrotik into the enterprise in the US.
I've been running wave2 at home since 7.7 with no complaints on a 4011 and an Audience, although I wasn't really using the 2.4ghz on the 4011 so I didn't really miss it.As someone have already mentioned - how realistic is wave2 on 40xx devices?
Upgraded a hAP ax2 to this version and now I keep getting a log message on every reboot with this:
"error while running customized default configuration script: no such item"
Any way to fix this? Downgrading back to 7.7 stable, didn't fix it.
5GHz Wi-Fi is “running” but clients fail to connect, working fine on 7.7 stable.
Yes. No. I do not want to netinstall as that's what I just did 5 days ago with this new box using 7.7. It's a lot of efforts for bugs that should be fixed by MikroTik.Have you tried reset with "no default configuration" option? Or even netinstall?
"DMVPN" is such a killer feature in the enterprise world. VyOS has support for it since years.
jan/20 14:16:47 system,error,critical error while running customized default configuration script: invalid internal item number
In the past this was often caused by renaming wlan interfaces in your local config (combined with programming errors in MikroTik's scripts).Upgraded a hAP ax2 to this version and now I keep getting a log message on every reboot with this:
"error while running customized default configuration script: no such item"
What is this limit? And is there a way to create a tmpfs with a sane default? Giving no size just fails now...*) disk - limit maximum TMPFS size;
[admin@MikroTik] > /disk/add type=tmpfs
failure: too much memory requested for tmpfs/ramdisk
Logging works fine unless you're trying some weird shit gathering data in splunk or whatever.
It may not be an importante function for you, but if you run MT in a larger scale, logging is important.
I suggest that you enable memory graphing and watch after a while what is the maximum amount of money consumed in your setup, and set the ramdisk so that it occupies less than the space that is always free.What is this limit? And is there a way to create a tmpfs with a sane default? Giving no size just fails now...*) disk - limit maximum TMPFS size;
I agree that it is a mess and needs improvement, but I fear that making any change to how it is now will cause an even bigger mess in existing installations than we have now.5 1/2 year and still no change in the logging prefix mess, that MT should look at.
No need to use bad words. Many people do use logging in various form, not only Splunk.Logging works fine unless you're trying some weird shit gathering data in splunk or whatever.
Yes, I know that. That's why I am not really happy with the change. I want to place backup files and exports on tmpfs, generated from scripts.Note that a ramdisk does not immediately occupy the space allocated to it. That happens only when it is filled with files.
new "rose-storage"…ARM, ARM64, Tile and x86
Any MIPS devices with more than 16MB of storage?
you can't fit that package on such little storage.
I just have been toying with it.Nobody noticed wave2 CAPSMAN GUI is here?
Thanks noted.Several comments:
Same problem here.Only read-only irq for xxv710 nic.may you elaborate little bit further?
*) x86 - fixed SR-IOV support for Intel X710 series NIC;
we use them and the only issue is the fact that we must use only auto about cpu irq interface queue. Is it something about?
yes and the strange think is that one queue is used a lot more the others, loading manly one core.Same problem here.Only read-only irq for xxv710 nic.may you elaborate little bit further?
*) x86 - fixed SR-IOV support for Intel X710 series NIC;
we use them and the only issue is the fact that we must use only auto about cpu irq interface queue. Is it something about?
nfs-share=...mount -t nfs server:/path /mountpoint
Thats properitary and requires a central controller somewhere, and can't co-exist with regular routing.in general called SD-WAN. This is implemented in v7 using ZeroTier.
I hope this is one of your jokes.the entire installation operates as a single AP MAC address on a single channel.
Nope, it's real.I hope this is one of your jokes.the entire installation operates as a single AP MAC address on a single channel.
What are the limitations for the CRS3xx devices that cause this version not to be recommended?
Important note!!!
Version is not recommended on CRS3xx devices.
Changes in this release:
example?Nope, it's real.
I hope this is one of your jokes.
No, really. There are manufacturers that offer a WiFi where there is only a single AP MAC for the controller, all APs listen on the same frequency and the same MAC, the traffic they receive is forwarded to the controller together with a signal strength value (RSSI), and the controller keeps the location of the strongest received signal so when they need to transmit data it can be sent via the AP where the signal is strongest.I hope this is one of your jokes.the entire installation operates as a single AP MAC address on a single channel.
which comes from acquisition of Meru NetworksFortinet is one https://www.fortinet.com/blog/industry- ... cell-wi-fi
It feels to me that, at least at this stage, it's meant more for router-to-router use, particularly for containers on diskless devices. I couldn't get a router to mount an iSCSI target from my TrueNAS box. I tried NFS and Samba from the 2116 to my Mac too. I couldn't get anything to work the way one would expect to another computer.Just tried rose-storage. Using 2 SATA-III SSDs inside RB1100, I can at least create a "mirrored disks" (RAID 1). So that part works!
...
3. Could NOT mount RouterOS ROSE disks using NFS Mac after using nfs-export=yes - tried various things: Finder using URL, Terminal using mount, using both :/ and :/raid1-part1 in path – nothing obvious seem to work and docs less helpful here.
4. The relationship with existing /ip/smb is unclear....
Yes - finally!! Confirmed fixed, Thanks!!*) container - fixed file ownership after system upgrade for containers running on internal disk;
*) container - fixed multiple container automatic startup on boot;
sudo mkdir /Volumes/2116/
sudo mount -t nfs -o vers=4,hard,bg,intr,resvport,rw 192.168.x.x:/nvme1 /Volumes/2116/
# add to /etc/auto_master if you don't already have it
/- auto_nfs -nobrowse,nosuid
# create /etc/auto_nfs, then add this line for Catalina and later; 2116 is the name of my share, change it to what you want
# nvme1 is my disk's name on the 2116
/System/Volumes/Data/../Data/Volumes/2116 -fstype=nfs,vers=4,hard,bg,intr,resvport,rw nfs://192.168.x.x/nvme1/
Flags: B - BLOCK-DEVICE; M, F - FORMATTING; p - PARTITION
Columns: SLOT, MODEL, SERIAL, INTERFACE, SIZE, FREE, FS, RAID-MASTER
# SLOT MODEL SERIAL INTERFACE SIZE FREE FS RAID
0 pcie1 SATA Controller PCIe 2x8 GT/s none
1 B pcie1-sata1 KINGSTON SH103S3 50026B73XXXXXXXX SATA 6.0 Gbps 240 057 409 536 none
2 BMp pcie1-sata1-part1 @512-240057409536 240 057 409 024 235 152 457 728 ext4 none
Now if I could only find a way to route those cables straight out a whole in the back and put the lid back on...
We get things like a disk manager, instead of some long awaited fixes in the basic functionality of a router.
Those modules present in "rose-storage" are all modules that required just to be enabled in the linux kernel, and the MikroTik team added some options to control them. Everything else that isn't upstreamed in the kernel yet isn't that easy to add.Nice
New storage and container fixes for routers
No core routing (BFD,EVPN, buggy VPN4) functionality for routers
Nice. No more words, sorry
Please, check 7.8beta2 bug with ipsec,error can't get private key SUP-105306
It help more people if you at least post the title/bug summary here. If everyone simply put the SUP#'s to check, the forum be way less useful.Please, check 7.8beta2 bug
SUP-105306
Those are too tall for the lid, and they'd hit the heat sink shroud.@sirbryan something like this [Startech Side Angle SATA cable]? Is there enough clearance for the cables to go off to the side?
By now it becomes more like "WE determine how long you can do without BFD!". It has been a "work in progress" for a year and a half, and thus it cannot be explained by "it is a complex thing to make" anymore.Bugs and things like BFD have no "enabled=yes" switch in RouterOS configuration files, sorry. It is easy to add some basic things, but there are complex things that need to be made from 0.
At this point the lack of BFD support is inexcusable. There is clear demand from users, it was previously implemented in RouterOS, it is implemented in competing stacks, and there are common situations for which no reasonable workaround exists. BFD is relatively simple compared with OSPF (though the v7 implementation of OSPF leaves a lot to be desired) so it is hard to understand why it is still missing and apparently receiving no attention whatsoever.Bugs and things like BFD have no "enabled=yes" switch in RouterOS configuration files, sorry. It is easy to add some basic things, but there are complex things that need to be made from 0.
Would be nice to have an alternative to hideously expensive Cisco/Juniper/Arista EVPN switches instead of relying on spanning-tree.No core routing (BFD,EVPN, buggy VPN4) functionality for routers
Sorry Normis, but writing nonsense like "there is no enabled=true for BFD" after 18+ months of claiming to work on it is just pathetic.Bugs and things like BFD have no "enabled=yes" switch in RouterOS configuration files, sorry
Agree. This issue is upgrade blocker cannot migrate due to it to ROS 7.pls fix BGP-VRF-VPNv4 - working with RR
I tend to agree. I can see that making RouterOS interface for mdadm/mkfs/iscsid/etc.. is low hanging fruit.. but why?What about having different ROS versions for SOHO boxes and serious routers? So one team can work on nerdy-stuff like containers and similar and another team can work on finally fixing and implementing basic routing stuff for CCR2x16 boxes?
<Raises hand> Um, I have six 2116's and just bought two more.</hand down> I also have a farm of NUCs, Mac Mini's, and Mac Pro 5,1's running ESXi or macOS, backed by TrueNAS storage.No one in their right mind buys a $995 router to play with containers. You buy a Mini-PC or some second hand Dell/HPE to play with virtualization and containers.
It is already fixed in v7.7 and v7.8betas
This will be my last post on this as it's getting off-topic, but ZeroTier is a pretty basic SD-WAN and is in no way equivalent to the capabilities, flexibility, and scalability of SD-WAN from vendors like Cisco-Viptela, Palo Alto-CloudGenix,VMware-VeloCloud, Fortinet SD-WAN, etc."DMVPN" is such a killer feature in the enterprise world. VyOS has support for it since years.
in general called SD-WAN. This is implemented in v7 using ZeroTier.
Can you please elaborate? What is fixed in v7.7 and v7.8 betas.It is already fixed in v7.7 and v7.8betas
No he can't, because that would be excessive quoting.Can you please elaborate? What is fixed in v7.7 and v7.8 betas.
This will be my last post on this as it's getting off-topic, but ZeroTier is a pretty basic SD-WAN and is in no way equivalent to the capabilities, flexibility, and scalability of SD-WAN from vendors like Cisco-Viptela, Palo Alto-CloudGenix,VMware-VeloCloud, Fortinet SD-WAN, etc.
MT will never play any role in any SDWAN unless the boxes can understand & detect (many) application and steer accordingly.This will be my last post on this as it's getting off-topic, but ZeroTier is a pretty basic SD-WAN and is in no way equivalent to the capabilities, flexibility, and scalability of SD-WAN from vendors like Cisco-Viptela, Palo Alto-CloudGenix,VMware-VeloCloud, Fortinet SD-WAN, etc.
mpvpn, meshvpn, sd-wan... same, same different name. I've used many of them thus the core functionality (l2/l3, end2end encryption, etc) and performance is more or less exactly the same.
What differs is deployment, administration and how well they jack into repective brands legacy system. I'd say the zt admin interface and api would fit any size as the client preparation process is the key factor that makes the difference in how well the deployment will work (as always).
Well, for one, MT doesn't drive the sd-wan market, they just adopted and implemented an existing open solution. Please feel free to elaborate what you mean by "unless the boxes can understand & detect (many) applications and steer accordingly." Btw, since this topic has nothing to do with v7.8Beta, please continue this discussion in another thread. I'll join you there in that case.MT will never play any role in any SDWAN unless the boxes can understand & detect (many) application and steer accordingly.
Everything mentioned in the changelogs.Can you please elaborate? What is fixed in v7.7 and v7.8 betas.It is already fixed in v7.7 and v7.8betas
despite being OT but SDWAN does not have anything to do with application detection.
MT will never play any role in any SDWAN unless the boxes can understand & detect (many) application and steer accordingly.
Of course it has! The selling point of SDWAN: it can replace great quality leased lines / MPLS by redundant cheap Internet uplinks. To achieve this you want to measure the quality of your cheap links and send important stuff over the better lines, while sending the rest of it over the worse lines.despite being OT but SDWAN does not have anything to do with application detection.
@mrz -said that RR has been fixedEverything mentioned in the changelogs
Can u explain more exactly what you have fixed for bpg vpn4?It is already fixed in v7.7 and v7.8betas
Sound really affordable to use Mac's as VMware hosts.<Raises hand> Um, I have six 2116's and just bought two more.</hand down> I also have a farm of NUCs, Mac Mini's, and Mac Pro 5,1's running ESXi or macOS, backed by TrueNAS storage.
For the same money I spent on my NUCs, the 2116 has more cores, similar RAM & storage, 40Gbps of connectivity into a Layer 3 switch with 16 ports, redundant power supplies, all in a rack mount case. And it can switch and route a heck of a lot more data.
Until containers came out for MikroTik, I didn't really have any experience with them, since most of my apps run on dedicated hosts or in VM's. Now I'm in the process of migrating much of what I ran on my Mac Pro ESXi host at home office onto my 2116, allowing me to save power and space and eventually decommission my Xeon servers.
With all the complaints about The DUDE dying or lacking in development (and the fact that it's stuck on Windows only), I can build or deploy the NMS of my choice, including sensors, on any number of RouterOS devices, with modern interfaces and customizable code.
Mikrotik started a different thread soliciting feedback on the topic a while back,: viewtopic.php?t=186352Please keep this forum topic strictly related to this particular RouterOS release.
That is not something a generic router can do. You need to buy a special box that has a maintenance contract to provide you with the dynamic information required for that.Believe me, while this is possible to achieve manually, you just want the box to "know" what o365 traffic is.
It was fix of issueCan u explain more exactly what you have fixed for bpg vpn4?It is already fixed in v7.7 and v7.8betas
It still have problem with best path calculation?
Thx
ROS 7 ibgp rr - ROS 6 ibgp .
Ros 6 establish connection and send update (MP_REACH_NLRI) with NLRI nexthop self ip (for example 10.29.193.134 )to ROS 7 RR
ROS 7 RR send back MP_REACH_NLRI with own self ip (for example 10.29.192.19 ) address and of course it also not care about propagate flag which also not right.
due to it ROS 6 reply MP_UNREACH_NLRI and ROS 7 reflect it back as well .
After it whole cycle repeat endless.
I attached wireshark capture , sipout from hub. and some screens which illustrate issue. 1 screen from ROS7 which show setting and 2 screens from ROS 6 where visible endless looped updates and export / import vrf rd.
The sarcasm does little to bolster your point. Those machines were made in 2009/2012 and ran through their useful macOS lifetime years ago.Sound really affordable to use Mac's as VMware hosts.
It can't be that heavy workloads if you can replace them with some ARM boxes with 16G of RAM.
:) And yet most of ASRs & ISRs of the biggest router vendor support it... search for "SD-WAN Application Intelligence Engine"That is not something a generic router can do. You need to buy a special box that has a maintenance contract to provide you with the dynamic information required for that.
What is a generic router? Even Mikrotik routers aren't "generic".That is not something a generic router can do. You need to buy a special box that has a maintenance contract to provide you with the dynamic information required for that.Believe me, while this is possible to achieve manually, you just want the box to "know" what o365 traffic is.
is that actually supported ?Still no hardware acceleration for OpenVPN tunnel and IPQ-6010 processor (hAP ax3)
What is a generic router? Even Mikrotik routers aren't "generic". FortiNet FortiGate firewalls can do NGFW, IPS, UTP, SD-WAN, BGP and OSPF. And SD-WAN doesn't require a license or maintenance contract. Cisco routers can do SD-WAN now too; it's an additional license but it's the same platforms. And the can both identify Office365. What does generic mean again?
Hi, can you elaborate a bit more on the scenario where you're facing this issue? I was considering to use CRS317 for a basic L3 routing with OSPF ...MT is currently missing a big point. Apart from routing features present in v6 that are not available on v7... they advertise the L3-HW that is BROKEN!
I have since 7.4 a support file, that is not fixed and they dindt give me any ETA for the fix.
We need every one hour to stop and restart the l3-hw engine on the CRS317. They are NOT READY for the isp business, but they ignore it.
SUP-95367 is still unfixed.
thank you for phrasing that out the way you did - that was the point why i mentioned SDWAN does not have to do anything with app-detection in it's basic implementation.
You are mixing up L7 firewalls and IDPS (intrusion detection and prevention systems) with routers. And the way you throw around acronyms, it's hard to tell if you even know the difference.
All types of application awareness are becoming increasingly difficult to achieve these days due to web-based applications and end-to-end encryption, you are left with just IP addresses and port numbers to play with. And for the same reason you can forget about "dynamic application routing" based on L7 filtering.
It will not be long before "destination port number" will be a constant that is always 443. We can get a new version of TCP that omits it.All types of application awareness are becoming increasingly difficult to achieve these days due to web-based applications and end-to-end encryption, you are left with just IP addresses and port numbers to play with. And for the same reason you can forget about "dynamic application routing" based on L7 filtering.
And most of these devices are a _LOT_ more expensive than even the CCR2216.:) And yet most of ASRs & ISRs of the biggest router vendor support it... search for "SD-WAN Application Intelligence Engine"That is not something a generic router can do. You need to buy a special box that has a maintenance contract to provide you with the dynamic information required for that.
Sincere apologies! I swear this is my last OT message in this thread!
And most of these devices are a _LOT_ more expensive than even the CCR2216... search for "SD-WAN Application Intelligence Engine"
What shit ? Please be more specific or nobody will be able to know what you mean.Hap ax3. Wifi 5Ghz speed same shit from v7.7...
@mrz
Can you guys please update your docs regarding MP-BGP? I still cannot get VPNv4 to work with either 7.7 or 7.8beta2, I can learn the routes from the route reflector, however traffic does not effectively flow. Either I am making some configuration mistake, or this still does not work.
Thanks in advance.
No, it's not.hAP ax3 wifi is very slow and unstable (but it's also on stable ROS):
I am very sorry, but it is.No, it's not.
Do VPLS tunnels work in your configuration?@mrz
Can you guys please update your docs regarding MP-BGP? I still cannot get VPNv4 to work with either 7.7 or 7.8beta2, I can learn the routes from the route reflector, however traffic does not effectively flow. Either I am making some configuration mistake, or this still does not work.
Thanks in advance.
For one, it doesn't help if you generalize.I am very sorry, but it is.No, it's not.
hAP ax2 & 3 have major stability and performance issues since their release for at least a few people.
Well made points....
Yes and no. We run IOS-XE on both ISR 1000 boxes and our 25/100G Catalyst. But the software image is about 800MB. Not the "within 16MB" principle that RouterOS has.Covering everything from hAP to CRS2x16 with one ROS seems to be a far stretch.
This is why I feel it might be worth to considering splitting up ROS into different versions for SOHO/CEP and large scale routers with different feature priorities.
Cisco IOS XE 16.12.1 introduces native Docker container support on certain models of Catalyst 9000 switches. This enables users to build and bring their own applications without additional packaging. Developers don’t have to reinvent the wheel by rewriting the applications every time there is an infrastructure change. Once packaged within Docker, the applications will work within any infrastructure that supports docker containers...
EOS offers the ability to write scripts and load applications directly onto the Linux operating system and to run these applications as guest VMs. Features of EOS extensibility include:
- Installation without modification of third-party software for Linux
- Scripting and Linux shell-level access for automation
- Programmable at all layers: Linux kernel, hardware forwarding tables, switch configuration and CLI, switch control plane as well as management layer
- Support for running Docker containers directly on the switch
You know what else also those Cisco/etc. routers can do? OSPF, BGP, BFD, and other basic routing protocols that we rightly expect a router to support. I do not doubt that there are legitimate use-cases for containers, but it remains secondary to core routing features -- especially on machines with support for L3 offloading and other routing/switching specific hardware features.For those wondering where MikroTik is going with all of this, especially CCR2000 and beyond...
From Cisco: https://www.cisco.com/c/en/us/products/ ... 42415.html (emphasis mine)
That is not necessarily the right thing, BGP signalled VPLS tunnels are created dynamically and there may be a lot of them in any given bridge. It is not clear to me how static names would even work in that case.also, we need static VPLS name , when we running them over BGP
I know, we have those at work. But using $9,000 switches for hosting containers is insane when you have a working VMware envoriment. The price difference GB/CPU per $ is astronomical. It only makes sense for a branch/edge site that needs a on-site agent of some kind (security, proxy, monitoring, authentication, etc.)From Cisco: https://www.cisco.com/c/en/us/products/ ... 42415.html (emphasis mine)
I applaud and welcome the fact that we also get it on the little ARM/ARM64 boxes. In the coming months, out of the hundreds of RouterOS devices I have installed in the field, more will be using newer features like Wireguard/ZeroTier and simple containers than MPLS/BGP/OSPF/etc.
Yes, with v7.7 and v7.8 speed slow... downgrade to v7.6 and OK.No, it's not.hAP ax3 wifi is very slow and unstable (but it's also on stable ROS):
With v7.6 5Ghz speed ~800Mbs, with v7.7 speed ~400Mbs using Iphone 12 pro maxWhat shit ? Please be more specific or nobody will be able to know what you mean.Hap ax3. Wifi 5Ghz speed same shit from v7.7...
I have mine since yesterday, upgraded to 7.8beta and I can easily get around 600mbs with Galaxy S20 (which I think is pretty decent).
On ac3 with wave2 it was low 500.
Where is your config and what hardware are your running on..With v7.6 5Ghz speed ~800Mbs, with v7.7 speed ~400Mbs using Iphone 12 pro max
hAP ax3Where is your config and what hardware are your running on..With v7.6 5Ghz speed ~800Mbs, with v7.7 speed ~400Mbs using Iphone 12 pro max
/ip/route/print count-only where
This is not the same! There isCli /routing/route
Gui ip -> Routes
/ip/route/
/routing/route/
What would really help is the ability to specify a VLAN for BGP-signalled VPLS. The need is somewhat niche, but I have found myself resorting to all kinds of brittle workarounds when it has come up.
Since updating to 7.8beta2 I'm having issues importing remote container image zabbix/zabbix-proxy-sqlite3:alpine-6.0-latest on my RB5009 (arm64). It was working fine on 7.7rc5.
Code: Select all16:43:05 container,info,debug importing remote image: zabbix/zabbix-proxy-sqlite3, tag: alpine-6.0-latest 16:43:05 system,info item added by cesar 16:43:07 container,info,debug error response getting manifests: 404 16:43:07 container,info,debug was unable to import, container 4a07240c-862b-4861-a16a-68605478ad54
After changing to zabbix/zabbix-proxy-sqlite3:alpine-6.0.12 it works fine again:
Code: Select all16:45:28 container,info,debug importing remote image: zabbix/zabbix-proxy-sqlite3, tag: alpine-6.0.12 16:45:28 system,info item added by cesar 16:45:31 container,info,debug getting layer sha256:6875df1f535433e5affe18ecfde9acb7950ab5f76887980ff06c5cdd48cf98f4 16:45:32 container,info,debug layer sha256:6875df1f535433e5affe18ecfde9acb7950ab5f76887980ff06c5cdd48cf98f4 downloaded 16:45:32 container,info,debug getting layer sha256:2068be5b412156c5bc2936aeb988446cb6ac458c4c408ac51b5143e9632073f0 16:45:33 container,info,debug layer sha256:2068be5b412156c5bc2936aeb988446cb6ac458c4c408ac51b5143e9632073f0 downloaded 16:45:33 container,info,debug getting layer sha256:35af6ce2b615d78f6617ef90fdbb0aef91a77c766594c28325a8e9e589d0e002 16:45:33 container,info,debug layer sha256:35af6ce2b615d78f6617ef90fdbb0aef91a77c766594c28325a8e9e589d0e002 downloaded 16:45:33 container,info,debug getting layer sha256:7becd6903f60f84a63358dbfbf033e34094e07d255085fe0d9a2fe48481e74b6 16:45:34 container,info,debug layer sha256:7becd6903f60f84a63358dbfbf033e34094e07d255085fe0d9a2fe48481e74b6 downloaded 16:45:34 container,info,debug getting layer sha256:21bb24f368b7ae4b135a1ef432a6379a54310c37e8a7b8d54d0260d7cd768f9d 16:45:35 container,info,debug layer sha256:21bb24f368b7ae4b135a1ef432a6379a54310c37e8a7b8d54d0260d7cd768f9d downloaded 16:45:35 container,info,debug getting layer sha256:9e1e869413aec50921ae70ba3b2098e56ab598bb6a26d2b0d5c697f7c433cb00 16:45:37 container,info,debug layer sha256:9e1e869413aec50921ae70ba3b2098e56ab598bb6a26d2b0d5c697f7c433cb00 downloaded 16:45:38 container,info,debug getting layer sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 16:45:38 container,info,debug layer sha256:4f4fb700ef54461cfa02571ae0db9a0dc1e0cdb5577484a6d75e68dc38e8acc1 downloaded 16:45:38 container,info,debug getting layer sha256:e4034d2118985bc524c23df0d8c998c604ee97aef464494c39111bf32ebd9335 16:45:39 container,info,debug layer sha256:e4034d2118985bc524c23df0d8c998c604ee97aef464494c39111bf32ebd9335 downloaded 16:45:39 container,info,debug import successful, container c3a27c76-186a-47bf-ace4-04fcff0790fd
zabbix/zabbix-proxy-sqlite3:alpine-6.0-latest was updated a few hours ago. Maybe something is wrong on Docker Hub? Or is it a bug in 7.8beta2?
I have the same problem with my own build, all newly pushed to docker hub are the same error. if you pull from other linux os and upload to routeros, it works.
I've opened a ticket (SUP-105409) but it was closed by MikroTik as "not our bug".
which indicates http error 404 (not found). That's a problem on remote side (github) and if things worked previously it may mean some necessary platform-specific files are missing.16:43:07 container,info,debug error response getting manifests: 404
my mistake :(belongs-to works fine for me! you should not use /ip/route/print but rather /routing/route/print.
Well, the "official" OpenVPN version 2.6.0 just released has dropped the support for static key, so that would be one possible reason for not bothering with it anymore...It would be great if OVPN would get static key support soon. Can you please tell me what's the problem implementing this?
Yes, but they write: static key mode (non-TLS) is no longer considered "good and secure enough" for today's requirements. Use TLS mode instead. If deploying a PKI CA is considered "too complicated", using --peer-fingerprint makes TLS mode about as easy as using --secret.”Well, the "official" OpenVPN version 2.6.0 just released has dropped the support for static key, so that would be one possible reason for not bothering with it anymore...
Ask three network engineers what "SDWAN" is and you will get at least three different answers.
MT will never play any role in any SDWAN unless the boxes can understand & detect (many) application and steer accordingly.
Sure, so long as you're using ARM. Three times now we have deployed Zerotier for a customer and *not* used CHR for the soft-router element because it doesn't support Zerotier.While I agree with many of the flaws, point #3 is generally referred to as SDWAN and is implemented in RoS using ZeroTier.
/ip dns
use-doh-server=https://dns-record.domain.tdl/dns-query verify-doh-cert=yes
/ip dns static
add address=public-ip name=dns-record.domain.tdl
When you want a simple no-certificate-hassle VPN, MikroTik offers more than enough alternatives: IPsec, SSTP, Wireguard, ...If deploying a PKI CA is considered "too complicated", using --peer-fingerprint makes TLS mode about as easy as using --secret.”
But Mikrotik does not support TLS authentication at all...
https://help.mikrotik.com/docs/display/ROS/OpenVPN (Limitations)
It's all about money: Our software solution provider says OpenVPN connection is free, IPsec connection is 410 Euro one-time-fee and 25 EUR monthly fee.When you want a simple no-certificate-hassle VPN, MikroTik offers more than enough alternatives: IPsec, SSTP, Wireguard, ...
When it is about VPN support, there is always somebody who asks for an option (or an entire protocol) that isn't supported...
"Solution provider" that offers OpenVPN only with static keys? That is pretty sad... they will not be happy finding that they need to point their users to obsolete software from now on, I guess...It's all about money: Our software solution provider says OpenVPN connection is free
......
I don't know if RouterOS can connect to it by now, it used to be not possible because it uses UDP only, and users with MikroTik routers get L2TP/IPsec or GRE tunnels instead.
I know that... there are many config parameters and also features that could make it fail. Compression, push of routes and other parameters, etc.RouterOS version 7 is now capable of using UDP OpenVPN connections. There are still some limitations, as it seems to use a proprietary implementation of OpenVPN.
me2, dns cache can't be freed though I flush dns cache since v7.7 stable. (tested on hap ac2 and RB4011)Not specific to 7.8b2, but seems to affect it. Starting from v7.7, it seems the dns resolver has a memory leak that eats up all the Mikrotik memory until the RouterOS crash and reboots. Not all are affected, but all affected confirmed to be providing DNS services for client devices.
If you're running 7.8 and providing DNS services for your connected networks, watch out for your box memory usage! This is really bad specially for the low-memory spec models, but seems to affect a variety of different models. Those with more RAM will just take longer to crash and (luckly) reboot.
viewtopic.php?t=192427#p979415
This does not seem to be working - I do not see this route either in the Winbox GUI or the CLI.*) route - fixed IPv6 default route presence when received from RA;
It works for me... maybe there is some other factor in your setup that influences it.This does not seem to be working - I do not see this route either in the Winbox GUI or the CLI.*) route - fixed IPv6 default route presence when received from RA;
SUP-105306*) certificate - improved certificate management, signing and storing processes;
Yeah, I figured out what the issue was.. I tried to build my own docker image and I couldn't get anything but 404 when I tried to pull the image to the Mikrotik. The issue is that the new Docker buildx uploads images in this OCI format. You'll also find that if you do `docker inspect <image name>` it will say the image isn't found because the inspect command also does not support this format. (funny right?) It's related to this issue: https://github.com/moby/moby/issues/43126 You can fix your own images by following this: https://github.com/docker/buildx/issues ... 1378538197I've opened a ticket (SUP-105409) but it was closed by MikroTik as "not our bug".
No wonder with error message
which indicates http error 404 (not found). That's a problem on remote side (github) and if things worked previously it may mean some necessary platform-specific files are missing.16:43:07 container,info,debug error response getting manifests: 404
Perhaps - but I just reset my hAP mini (which I only use for testing) to no-default-configuration, logged into it and disabled "IPv6 forward" so that it gets an address from the RA. The address appears, but default route does not, but the default route is actually there because it can ping the IPv6 address for ipv6.google.com.It works for me... maybe there is some other factor in your setup that influences it.
I did not have IPv6 forwarding enabled on my test router, I use IPv6 only for a potential tunnel there, but after enabling it it still works.It doesn't work on my hAP ac either. If it is some other factor in the setup, it must be something about the RA itself that it doesn't like, preventing it from installing the default route for some reason.
*) console - improved ":execute" command to output a string when a file is not specified;
[admin@MikroTik] > :global x [:execute ":put \"test\""]
[admin@MikroTik] > put $x
test
:global z [:execute ":while (1) do={:put here; :delay 1s}"]
I wouldn't say explained. :execute has worked for years the same way – it does NOT wait for a return value, and that will break some existing scripts people use.The change is explained in this video:
https://youtu.be/xYLYRmpM-Zo?t=217
What's new in 7.8beta3 (2023-Feb-01 16:10):
Important note!!!
Version is not recommended on CRS3xx devices.
Changes in this release:
*) certificate - fixed PBES2 certificate import;
*) certificate - improved multiple certificate import process;
/certificate/add name="r1-ca" common-name="r1-ca" subject-alt-name="email:r1-ca" key-size=prime256v1 key-usage=key-cert-sign,crl-sign
:do {/certificate/sign [find name=r1-ca] name=r1-ca} on-error={:delay 3}
/certificate/add name="r1" common-name="192.168.2.14" subject-alt-name="IP:192.168.2.14" key-size=prime256v1 key-usage=digital-signature,content-commitment,key-encipherment,key-agreement,tls-server
:do {/certificate/sign [find name=r1] ca=r1-ca name=r1} on-error={:delay 3}
/certificate/add name="r1-r2" common-name="r1-r2" subject-alt-name="email:r1-r2" key-size=prime256v1 key-usage=digital-signature,key-encipherment,data-encipherment,key-agreement,tls-client
:do {/certificate/sign [find name=r1-r2] ca=r1-ca name=r1-r2} on-error={:delay 3}
:delay 2
/certificate/export-certificate r1-ca file-name=r1-ca
/certificate/export-certificate r1 file-name=r1
/certificate/export-certificate r1-r2 file-name=r1-r2 type=pkcs12 export-passphrase=passphrase
/ip/pool/add name=r1-r2 ranges=192.168.1.2
/ip/ipsec/mode-config/add address-pool=r1-r2 address-prefix-length=32 name=r1-r2 split-include=0.0.0.0/0 system-dns=no
/ip/ipsec/policy/group/add name=group1
/ip/ipsec/profile/add dh-group=ecp256 enc-algorithm=aes-256 hash-algorithm=sha256 name=profile1 prf-algorithm=sha256 proposal-check=strict
/ip/ipsec/peer/add exchange-mode=ike2 local-address=192.168.2.14 name=peer1 passive=yes profile=profile1
/ip/ipsec/proposal/add auth-algorithms=sha256 enc-algorithms=aes-256-cbc,aes-256-gcm lifetime=8h name=proposal1 pfs-group=ecp256
/ip/ipsec/identity/add auth-method=digital-signature certificate=r1 generate-policy=port-strict match-by=certificate mode-config=r1-r2 peer=peer1 policy-template-group=group1 remote-certificate=r1-r2
/ip/ipsec/policy/add dst-address=192.168.1.0/24 group=group1 proposal=proposal1 src-address=0.0.0.0/0 template=yes
/certificate/import file-name="r1-ca.crt" name="r1-ca" passphrase=""
/certificate/import file-name="r1.crt" name="r1" passphrase=""
/certificate/import file-name="r1-r2.p12" name="r1-r2" passphrase="passphrase"
:delay 2
/ip/ipsec/mode-config/add name=cfg1 responder=no
/ip/ipsec/policy/group/add name=group1
/ip/ipsec/profile/add dh-group=ecp256 enc-algorithm=aes-256 hash-algorithm=sha256 name=profile1 prf-algorithm=sha256 proposal-check=strict
/ip/ipsec/peer/add address=192.168.2.14/32 exchange-mode=ike2 name=peer1 profile=profile1
/ip/ipsec/proposal/add auth-algorithms="" enc-algorithms=aes-256-gcm lifetime=8h name=proposal1 pfs-group=ecp256
/ip/ipsec/identity/add auth-method=digital-signature certificate=r1-r2 generate-policy=port-strict match-by=certificate mode-config=cfg1 my-id=dn peer=peer1 policy-template-group=group1 remote-certificate=r1
/ip/ipsec/policy/add dst-address=0.0.0.0/0 group=group1 proposal=proposal1 src-address=0.0.0.0/0 template=yes
[admin@MikroTik] > /log/print
17:56:50 system,info crossfig will upgrade version 6 configuration
17:56:50 system,info router rebooted
17:56:56 dhcp,info dhcp-client on ether1 got IP address 192.168.2.14
17:57:26 system,info,account user admin logged in from 192.168.2.12 via winbox
17:57:35 system,info,account user admin logged in from 192.168.2.12 via local
17:57:55 certificate,info generated CA certificate: r1-ca
17:57:55 certificate,info generated certificate 58D11DB0B6FC086E:192.168.2.14::::::IP:192.168.2.14 ec-curve:prime256v1 usage:80000017 valid:365 for CA r1-ca
17:57:55 certificate,info generated certificate 635C8FE1F8067C04:r1-r2::::::email:r1-r2 ec-curve:prime256v1 usage:4000001d valid:365 for CA r1-ca
17:57:57 system,info pool r1-r2 added by admin
17:57:57 system,info ipsec modecfg r1-r2 added by admin
17:57:57 system,info ipsec policy group added by admin
17:57:57 system,info peer proposal profile1 added by admin
17:57:57 system,info ipsec peer peer1 added by admin
17:57:57 system,info ipsec proposal proposal1 added by admin
17:57:57 system,info ipsec identity added by admin
17:57:57 system,info ipsec policy added by admin
17:58:17 ipsec,info new ike2 SA (R): peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:3c1d6cb395cc01d2:6d092be31bed4e80
17:58:17 ipsec,error got fatal error: AUTHENTICATION_FAILED
17:58:17 ipsec,info killing ike2 SA: peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:3c1d6cb395cc01d2:6d092be31bed4e80
17:58:27 ipsec,info new ike2 SA (R): peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:60f9761f291a8e80:7f0c0161d5d1c77b
17:58:27 ipsec,error got fatal error: AUTHENTICATION_FAILED
17:58:27 ipsec,info killing ike2 SA: peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:60f9761f291a8e80:7f0c0161d5d1c77b
17:58:37 ipsec,info new ike2 SA (R): peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:e7a61a9079b43bde:2554939c1b6bef4f
17:58:37 ipsec,error got fatal error: AUTHENTICATION_FAILED
17:58:37 ipsec,info killing ike2 SA: peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:e7a61a9079b43bde:2554939c1b6bef4f
17:58:47 ipsec,info new ike2 SA (R): peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:f8f8ffcf6c778b67:f72f141417a39316
17:58:47 ipsec,error got fatal error: AUTHENTICATION_FAILED
17:58:47 ipsec,info killing ike2 SA: peer1 192.168.2.14[4500]-192.168.2.15[4500] spi:f8f8ffcf6c778b67:f72f141417a39316
[admin@MikroTik] > /log/print
17:56:58 system,info crossfig will upgrade version 6 configuration
17:56:58 system,info router rebooted
17:57:04 dhcp,info dhcp-client on ether1 got IP address 192.168.2.15
17:57:29 system,info,account user admin logged in from 192.168.2.12 via winbox
17:57:36 system,info,account user admin logged in from 192.168.2.12 via local
17:58:17 system,info ipsec modecfg cfg1 added by admin
17:58:17 system,info ipsec policy group added by admin
17:58:17 system,info peer proposal profile1 added by admin
17:58:17 system,info ipsec peer peer1 added by admin
17:58:17 system,info ipsec proposal proposal1 added by admin
17:58:17 system,info ipsec identity added by admin
17:58:17 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:6d092be31bed4e80:3c1d6cb395cc01d2
17:58:17 ipsec,error can't get private key
17:58:17 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:6d092be31bed4e80:3c1d6cb395cc01d2
17:58:21 system,info ipsec policy added by admin
17:58:27 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:7f0c0161d5d1c77b:60f9761f291a8e80
17:58:27 ipsec,error can't get private key
17:58:27 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:7f0c0161d5d1c77b:60f9761f291a8e80
17:58:37 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:2554939c1b6bef4f:e7a61a9079b43bde
17:58:37 ipsec,error can't get private key
17:58:37 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:2554939c1b6bef4f:e7a61a9079b43bde
17:58:46 system,info,account user admin logged in from 192.168.2.12 via local
17:58:47 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:f72f141417a39316:f8f8ffcf6c778b67
17:58:47 ipsec,error can't get private key
17:58:47 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:f72f141417a39316:f8f8ffcf6c778b67
17:58:57 ipsec,info new ike2 SA (I): peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:092169a7e0132082:de8a29cf2fbc6a16
17:58:57 ipsec,error can't get private key
17:58:57 ipsec,info killing ike2 SA: peer1 192.168.2.15[4500]-192.168.2.14[4500] spi:092169a7e0132082:de8a29cf2fbc6a16
depth0cert - Changelog entries that you reder to are related to the "/ertificate import" command, not to the problem tha tyou refer to.
Many thanks :)*) zeroter - fixed routes after VRF change;
.............................*) console - improved ":execute" command to output a string when a file is not specified;
ROSE is pretty cool. I tried Btrfs/ext4 but they all do not allow faster SFTP transfers than 14Mbyte/s on a CCR2216 with NVMe. The interface says PCIe 4x8 GT/s and 31.5 Gbps.dksoft - We will see about that... New packages are included in stable versions when we can consider them "stable". At the moment we can not promise that ROSE-storage will be included in v7.8, but we will do our best in order to make that happen. Of course, even if it will not be released with 7.8, it will be back in 7.9beta releases.
................................
{
:put "OK"
:put [:execute ":resolve notexist.sht" file=/useless_change_that_broke_the_balls]
:put "OK"
}
>"OK"
>*24
>"OK"
{ :put "OK" :put [:execute ":resolve notexist.sht" return-errors=yes] # return-errors can be no (and must be the default for keep the integrity of what already exist) or yes (the new wanted feature) :put "OK" } >OK >failure: dns name does not exist >OKHowever this addition IS ABSOLUTELY USEFUL... I already have some ideas in mind...
Perhaps an "as-string" or "wait=yes" or whatever option to trigger it. Or it be better still if somehow preserved types on the return (e.g. "as-value"). More say existing syntax, while more complex, would have solve the telegram script's need & if y'all were going to muck with scripting to make things easier, way better targets than changing :execute.But it's how it's implemented, which destroys all existing scripts (which use ":execute", of course), that is bulls–t.
*) ike2 - fixed active SA flush on responder after an unsuccessful peer connection attempt;
Another big one is "count-only". You can print count-only, but you cannot "get" it. And when you print count-only you cannot use "as-value".I'd say MT is pretty consistent with "as-value" (other than here), with only "/[cmd] get" implying as-value (since "print" is what you use at CLI instead of "get"). The only example where an as-value is missing that I recall is /tool/snmp-get (which I reported 2 years ago and remains unfixed).
Fair enough, but [:len [find]] kinda does that for you. (And fixing the performance of the [find] in the first place ;) )Another big one is "count-only". You can print count-only, but you cannot "get" it.
Do you really think that when I want to know the number of routes for a certain "where" query, I need to fetch them all into an array and count the number of elements?Fair enough, but [:len [find]] kinda does that for you. (And fixing the performance of the [find] in the first place ;) )Another big one is "count-only". You can print count-only, but you cannot "get" it.
I believe you, but more odd that count-only be quicker, it has to do some enumeration. Anyway, certainly not arguing against some "print count-only as-value" – as that doesn't break anything in scriptingDo you really think that when I want to know the number of routes for a certain "where" query, I need to fetch them all into an array and count the number of elements?
That is ridiculous, isn't it?
Hun, i think i'm affected by this (on CRS317 running V6.49)*) bridge - fixed adding disabled MSTI;
10:14:30 interface,info ether5 link down
10:14:33 interface,info ether5 link up (speed 1G, full duplex)
10:15:35 interface,info sfp-sfpplus1 link down
10:15:35 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:16:34 interface,info sfp-sfpplus1 link down
10:16:34 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:16:42 interface,info sfp-sfpplus1 link down
10:16:42 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:17:00 interface,info sfp-sfpplus1 link down
10:17:01 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:17:29 interface,info sfp-sfpplus1 link down
10:17:29 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:18:17 interface,info sfp-sfpplus1 link down
10:18:17 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:18:35 poe-out,debug ether5 detected poe-out status: disabled
10:18:35 interface,info ether5 link down
10:18:45 poe-out,debug ether5 detected poe-out status: wait_for_load
10:18:45 poe-out,debug ether5 detected poe-out status: on
10:18:53 interface,info ether5 link up (speed 1G, full duplex)
10:19:01 interface,info sfp-sfpplus1 link down
10:19:01 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:19:03 interface,info ether5 link down
10:19:16 interface,info ether5 link up (speed 1G, full duplex)
10:19:18 interface,info sfp-sfpplus1 link down
10:19:18 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:19:42 interface,info sfp-sfpplus1 link down
10:19:43 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:20:00 interface,info sfp-sfpplus1 link down
10:20:00 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:21:18 interface,info ether1 link down
10:21:20 interface,info ether1 link up (speed 2.5G, full duplex)
10:21:32 interface,info sfp-sfpplus1 link down
10:21:32 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:21:33 interface,info sfp-sfpplus1 link down
10:21:33 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:21:37 interface,info sfp-sfpplus1 link down
10:21:37 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:21:47 interface,info sfp-sfpplus1 link down
10:21:47 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:21:51 interface,info ether1 link down
10:21:52 interface,info ether1 link up (speed 2.5G, full duplex)
10:21:55 interface,info ether1 link down
10:22:01 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:03 interface,info ether1 link down
10:22:04 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:07 interface,info ether1 link down
10:22:09 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:10 interface,info sfp-sfpplus1 link down
10:22:10 interface,info sfp-sfpplus1 link up (speed 10G, full duplex)
10:22:22 interface,info ether1 link down
10:22:24 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:38 interface,info ether1 link down
10:22:40 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:43 interface,info ether1 link down
10:22:44 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:48 interface,info ether1 link down
10:22:50 interface,info ether1 link up (speed 2.5G, full duplex)
10:22:53 interface,info ether1 link down
10:22:55 interface,info ether1 link up (speed 2.5G, full duplex)
10:23:07 interface,info ether1 link down
10:23:13 interface,info ether1 link up (speed 2.5G, full duplex)
10:23:15 interface,info ether1 link down
10:23:16 interface,info ether1 link up (speed 2.5G, full duplex)
10:26:07 interface,info ether1 link down
10:26:10 interface,info ether1 link up (speed 2.5G, full duplex)
can someone please further describe what are the impacts those releases are not recommended on CRS3xx devices?What's new in 7.8beta3 (2023-Feb-01 16:10):
Important note!!!
Version is not recommended on CRS3xx devices.
What's new in 7.8beta3 (2023-Feb-01 16:10):
Important note!!!
Version is not recommended on CRS3xx devices.
Changes in this release:
*) certificate - fixed PBES2 certificate import;
*) certificate - improved multiple certificate import process;
[admin@MikroTik] > /certificate add common-name=r1-ca days-valid=3652 key-size=prime256v1 key-usage=key-cert-sign,crl-sign name=r1-ca subject-alt-name=email:r1-ca
[admin@MikroTik] > :do {/certificate sign [find name=r1-ca] name=r1-ca} on-error={:delay 3}
progress: done
[admin@MikroTik] > /certificate export-certificate r1-ca file-name=r1-ca export-passphrase=passphrase type=pem
[admin@MikroTik] > /certificate remove r1-ca
[admin@MikroTik] > /certificate import file-name=r1-ca.crt name=r1-ca passphrase=passphrase
certificates-imported: 1
private-keys-imported: 0
files-imported: 0
decryption-failures: 0
keys-with-no-certificate: 0
[admin@MikroTik] > /certificate import file-name=r1-ca.key name=r1-ca passphrase=passphrase
certificates-imported: 0
private-keys-imported: 1
files-imported: 1
decryption-failures: 0
keys-with-no-certificate: 0
[admin@MikroTik] > /certificate add common-name=r1-ca days-valid=3652 key-size=prime256v1 key-usage=key-cert-sign,crl-sign name=r1-ca subject-alt-name=email:r1-ca
[admin@MikroTik] > :do {/certificate sign [find name=r1-ca] name=r1-ca} on-error={:delay 3}
progress: done
[admin@MikroTik] > /certificate export-certificate r1-ca file-name=r1-ca export-passphrase=passphrase type=pem
[admin@MikroTik] > /certificate remove r1-ca
[admin@MikroTik] > /certificate import file-name=r1-ca.crt name=r1-ca passphrase=passphrase
certificates-imported: 1
private-keys-imported: 0
files-imported: 0
decryption-failures: 0
keys-with-no-certificate: 0
[admin@MikroTik] > /certificate import file-name=r1-ca.key name=r1-ca passphrase=passphrase
certificates-imported: 0
private-keys-imported: 1
files-imported: 1
decryption-failures: 0
keys-with-no-certificate: 0
[admin@MikroTik] > /certificate add common-name=r1-ca days-valid=3652 key-size=prime256v1 key-usage=key-cert-sign,crl-sign name=r1-ca subject-alt-name=email:r1-ca
[admin@MikroTik] > :do {/certificate sign [find name=r1-ca] name=r1-ca} on-error={:delay 3}
progress: done
[admin@MikroTik] > /certificate export-certificate r1-ca file-name=r1-ca export-passphrase=passphrase type=pem
[admin@MikroTik] > /certificate remove r1-ca
[admin@MikroTik] > /certificate import file-name=r1-ca.crt name=r1-ca passphrase=passphrase
certificates-imported: 1
private-keys-imported: 0
files-imported: 0
decryption-failures: 0
keys-with-no-certificate: 0
[admin@MikroTik] > /certificate import file-name=r1-ca.key name=r1-ca passphrase=passphrase
certificates-imported: 0
private-keys-imported: 0
files-imported: 0
decryption-failures: 0
keys-with-no-certificate: 0
What's new in 7.8beta3 (2023-Feb-01 16:10):
Important note!!!
Version is not recommended on CRS3xx devices.
Changes in this release:
I have flapping ports also, but not POE related. I use 7.8beta3 on my Rb450G4 and ax3 connected to it (also cap ac). ax3 reboots whit kernel failure quite often.After upgrade to 7.8beta3 RB5009UPr+ PoE issue happen again. fix on 7.8alpha227
please check ticket SUP-105042
on ether1-2.5G is UBNT U6-Enterprise
on ether5-1G is UBNT U6-LR
Code: Select all10:14:30 interface,info ether5 link down 10:14:33 interface,info ether5 link up (speed 1G, full duplex) 10:15:35 interface,info sfp-sfpplus1 link down 10:15:35 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:16:34 interface,info sfp-sfpplus1 link down 10:16:34 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:16:42 interface,info sfp-sfpplus1 link down 10:16:42 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:17:00 interface,info sfp-sfpplus1 link down 10:17:01 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:17:29 interface,info sfp-sfpplus1 link down 10:17:29 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:18:17 interface,info sfp-sfpplus1 link down 10:18:17 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:18:35 poe-out,debug ether5 detected poe-out status: disabled 10:18:35 interface,info ether5 link down 10:18:45 poe-out,debug ether5 detected poe-out status: wait_for_load 10:18:45 poe-out,debug ether5 detected poe-out status: on 10:18:53 interface,info ether5 link up (speed 1G, full duplex) 10:19:01 interface,info sfp-sfpplus1 link down 10:19:01 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:19:03 interface,info ether5 link down 10:19:16 interface,info ether5 link up (speed 1G, full duplex) 10:19:18 interface,info sfp-sfpplus1 link down 10:19:18 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:19:42 interface,info sfp-sfpplus1 link down 10:19:43 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:20:00 interface,info sfp-sfpplus1 link down 10:20:00 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:18 interface,info ether1 link down 10:21:20 interface,info ether1 link up (speed 2.5G, full duplex) 10:21:32 interface,info sfp-sfpplus1 link down 10:21:32 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:33 interface,info sfp-sfpplus1 link down 10:21:33 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:37 interface,info sfp-sfpplus1 link down 10:21:37 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:47 interface,info sfp-sfpplus1 link down 10:21:47 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:51 interface,info ether1 link down 10:21:52 interface,info ether1 link up (speed 2.5G, full duplex) 10:21:55 interface,info ether1 link down 10:22:01 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:03 interface,info ether1 link down 10:22:04 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:07 interface,info ether1 link down 10:22:09 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:10 interface,info sfp-sfpplus1 link down 10:22:10 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:22:22 interface,info ether1 link down 10:22:24 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:38 interface,info ether1 link down 10:22:40 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:43 interface,info ether1 link down 10:22:44 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:48 interface,info ether1 link down 10:22:50 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:53 interface,info ether1 link down 10:22:55 interface,info ether1 link up (speed 2.5G, full duplex) 10:23:07 interface,info ether1 link down 10:23:13 interface,info ether1 link up (speed 2.5G, full duplex) 10:23:15 interface,info ether1 link down 10:23:16 interface,info ether1 link up (speed 2.5G, full duplex) 10:26:07 interface,info ether1 link down 10:26:10 interface,info ether1 link up (speed 2.5G, full duplex)
07:12:02 interface,info ether1 link up (speed 1G, full duplex)
09:07:54 interface,info ether1 link down
09:08:08 interface,info ether1 link up (speed 100M, full duplex)
09:08:46 interface,info ether1 link down
09:09:00 interface,info ether1 link up (speed 100M, full duplex)
Not sure what do you mean by "get", if you mena to use the value in scripts you can certainly do itAnother big one is "count-only". You can print count-only, but you cannot "get" it. And when you print count-only you cannot use "as-value".
[admin@3C22-atombumba] /ip/route> :global a [print count-only ]
36
[admin@3C22-atombumba] /ip/route> :put $a
36
The problem is the first printing of "36". That should not happen. I want to "get" the value in a variable, NOT "print it on the terminal".Not sure what do you mean by "get", if you mena to use the value in scripts you can certainly do itAnother big one is "count-only". You can print count-only, but you cannot "get" it. And when you print count-only you cannot use "as-value".
Code: Select all[admin@3C22-atombumba] /ip/route> :global a [print count-only ] 36 [admin@3C22-atombumba] /ip/route> :put $a 36
I want to make a script that I can start from the terminal and that collects some values like that and prints them in a formatted table.The point is that you can use it in the scripts.
This was about rsync support in ROSE-package as statet out in the online documentation.dksoft - We will see about that...
SFTP is slow single connection protocol, run multiple instances parallel to get faster results.What is the reason for that slow transfer?
This make some sense to me. I'm just note sure it comes up in a lot of cases... I still think the problem with your "print count-only" is that RouterOS should respect an "as-value" which could prevent the output in your formatted table. (Or [find] should be closer in performance ;))I think the :set variable [command] construct should run any command and put the output of that command in the variable, and not on the terminal.The point is that you can use it in the scripts.
I am unable to recreate this on my 5G or LTE12, are you updating both routerboard and RouterOS?After upgrading Chateau LTE12 (RBD53G-5HacD2HnD) and Chateau 5G devices from 7.8beta2 to 7.8beta3 the LEDs indicating modem signal strength stop working (no LEDs for signal strength 1-5 are working, other LEDs are working as expected), rebooting device doesn't fix the issue.
It would be the same as the $(command) construct in e.g. bash. It is handy in a lot of cases!This make some sense to me. I'm just note sure it comes up in a lot of cases...I think the :set variable [command] construct should run any command and put the output of that command in the variable, and not on the terminal.
All of what you write is about low level scripting languages. In bash you have just variable substitution ($(command)), in- and output redirection (|, <, >, ...) and numerical return code (available with $? or when using && and ||).It would be the same as the $(command) construct in e.g. bash. It is handy in a lot of cases!
What I still do not understand is how it would be design to have [command] return a value for use in a variable assignment, but AT THE SAME TIME still output to the terminal!
Output redirection (into a file, a variable, a pipe, whatever) always REPLACES the default output to the terminal, in any other system.
To have both the redirected output and the plain terminal output, one would expect a specific duplication tool like "tee" to be used.
The docs are ahead of the release, that's a welcomed first!:execute
Execute the script in the background.
The result can be written in the file by setting a "file" parameter or printed to the CLI by setting "as-string".
When using "as-string" parameter executed script is blocking (not executed in the background).
I am ok with that. Just more worried that 7.7 would be the last update for them.Because it is a beta version and they know already in advance it doesn't work on that platform.
It needs more work.
Simple.
Would you prefer it otherwise ?
The docs are ahead of the release, that's a welcomed first!
it won't. why would it?
I am ok with that. Just more worried that 7.7 would be the last update for them.
:execute change was a mistake for this beta and will be fixed in the next beta.And still a -1 to the :execute changes here.
Please give me a help: All I find is this page: https://help.mikrotik.com/docs/display/ROS/ROSE-storageFor rsync - please read rose manual page.
:execute change was a mistake for this beta and will be fixed in the next beta.
Wait! $(command) is NOT variable substitution! It means: run "command", catch its output, and return that as a value. It was before known as `command`.All of what you write is about low level scripting languages. In bash you have just variable substitution ($(command))
I agree, but that is not what I am asking for. What I am asking for is that ANY command that outputs text to the terminal can be run in a [command] construct and that will return the output in a string form (maybe an array when it is multi-line) and it will NOT print the same thing on the terminal. Of course there would still be the possibility of returning structured data in some situations using the strangely named as-value option (why not as-array?).The scripting in RouterOS is more like higher level scripting or programming languages here. There you can make a function return complex variables, data structures or objects. Please do not break this.
I think there are unexpected and unreasonable limitations in the expression evaluator (you cannot write everything you want as a single expression, at some point you need to split it in multiple steps or it will not work correctly), and also the diagnostics and debug info in such cases is really problematic.There are some rough edges in RouterOS scripting, but most of this is really well done.
And another issue I just noticed on beta 3 my Dyndns script cant finish updating IP, it doesn't come to part where it writes to disk, but If I manually run script instead scheduler it finishes.
/system scheduler add interval=5m name=dyndns on-event="/system script run dyndns" policy=ftp,reboot,read,write,policy,test,password,sniff,sensitive,romon start-date=may/04/2020 start-time=00:00:00
/system script add dont-require-permissions=no name=dyndns owner=admin policy=ftp,reboot,read,write,policy,test,password,sniff,sensitive,romon source="
\n:local myip [/ip/address/get [find where interface=pppoe1] address]\r\
\n:set myip [:pick \$myip 0 [:find \$myip \"/\"]]\r\
\n\r\
\n/tool fetch keep-result=no url=(\"http://freedns.afraid.org/dynamic/update.php\\\?abcdREMOVEDefghijk=&address=\$myip\")\r\
\n/tool fetch keep-result=no url=(\"https://www.duckdns.org/update\?domains=REMOVED&token=abcdREMOVEDefghijk&ip=\$myip\")"
What still surprises me is that that luxurious ATLASSIAN documentation system does not support documentation tied to a product release version...The docs are ahead of the release, that's a welcomed first!
What still surprises me is that that luxurious ATLASSIAN documentation system does not support documentation tied to a product release version...
Same here, I changed toI noticed same issue here. My dyndns script doesn't update anymore from on scheduler (although the "Last Time Started" counter for the script does suggest the script runs). If I execute the script manually it works.And another issue I just noticed on beta 3 my Dyndns script cant finish updating IP, it doesn't come to part where it writes to disk, but If I manually run script instead scheduler it finishes.
/tool fetch url="..." as-value output=user
Mikrotik have a policy of supporting any given model for 5 years after it's discontinued. At least had, and I don't think it changed. So...I am ok with that. Just more worried that 7.7 would be the last update for them.
:execute change was a mistake for this beta and will be fixed in the next beta.And still a -1 to the :execute changes here.
Could you share full script as you got working now if its not a problem ?Same here, I changed to
I noticed same issue here. My dyndns script doesn't update anymore from on scheduler (although the "Last Time Started" counter for the script does suggest the script runs). If I execute the script manually it works.and things worked again.Code: Select all/tool fetch url="..." as-value output=user
Why if its 100% related to beta 3 update, it needs to be adressed and known why it happens, this script worked untouched for 5 + years for me on 100+ routers, its first time stoped working on v7.8 beta.3.Open separate topic, or go to one of the thousand topic about dyndns & similar that already exist
I am running 7.8b3 on HAP AC3 without any issues.7.8beta2 is more stable than 7.8beta3
tested on hex s and hap ac3, same behavior.
In such cases, export the config, netinstall the device, and import the config from the export file.7.8beta2 is more stable than 7.8beta3
had problem with device freezes with 7.7, tried to use 7.8beta2 problem was resolved, now after updating to 7.8beta3 device again started to freeze or random reboot with kernel failure...
Even sigma?@holvoetn
Wasted time, people instantly install anything new, they install it right away, they don't care if it's alpha, beta or omega, and often don't even read the release notes...
It was POE port related. ax3 in my network topology is connected to rb450g4 on port 5 (that is POE). I am powering device whit mains power, but still POE ports did their thing. As result ax3 crashed whit kernel failure, rb450g4 just flapped ports, 4 bridge ports. I disabled POE feature on both devices and issue stopped. ax3 have no downtime and rb450g4 port are up for 4 days. On 7.7 all was fine.I have flapping ports also, but not POE related. I use 7.8beta3 on my Rb450G4 and ax3 connected to it (also cap ac). ax3 reboots whit kernel failure quite often.After upgrade to 7.8beta3 RB5009UPr+ PoE issue happen again. fix on 7.8alpha227
please check ticket SUP-105042
on ether1-2.5G is UBNT U6-Enterprise
on ether5-1G is UBNT U6-LR
Code: Select all10:14:30 interface,info ether5 link down 10:14:33 interface,info ether5 link up (speed 1G, full duplex) 10:15:35 interface,info sfp-sfpplus1 link down 10:15:35 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:16:34 interface,info sfp-sfpplus1 link down 10:16:34 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:16:42 interface,info sfp-sfpplus1 link down 10:16:42 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:17:00 interface,info sfp-sfpplus1 link down 10:17:01 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:17:29 interface,info sfp-sfpplus1 link down 10:17:29 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:18:17 interface,info sfp-sfpplus1 link down 10:18:17 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:18:35 poe-out,debug ether5 detected poe-out status: disabled 10:18:35 interface,info ether5 link down 10:18:45 poe-out,debug ether5 detected poe-out status: wait_for_load 10:18:45 poe-out,debug ether5 detected poe-out status: on 10:18:53 interface,info ether5 link up (speed 1G, full duplex) 10:19:01 interface,info sfp-sfpplus1 link down 10:19:01 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:19:03 interface,info ether5 link down 10:19:16 interface,info ether5 link up (speed 1G, full duplex) 10:19:18 interface,info sfp-sfpplus1 link down 10:19:18 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:19:42 interface,info sfp-sfpplus1 link down 10:19:43 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:20:00 interface,info sfp-sfpplus1 link down 10:20:00 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:18 interface,info ether1 link down 10:21:20 interface,info ether1 link up (speed 2.5G, full duplex) 10:21:32 interface,info sfp-sfpplus1 link down 10:21:32 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:33 interface,info sfp-sfpplus1 link down 10:21:33 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:37 interface,info sfp-sfpplus1 link down 10:21:37 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:47 interface,info sfp-sfpplus1 link down 10:21:47 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:21:51 interface,info ether1 link down 10:21:52 interface,info ether1 link up (speed 2.5G, full duplex) 10:21:55 interface,info ether1 link down 10:22:01 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:03 interface,info ether1 link down 10:22:04 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:07 interface,info ether1 link down 10:22:09 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:10 interface,info sfp-sfpplus1 link down 10:22:10 interface,info sfp-sfpplus1 link up (speed 10G, full duplex) 10:22:22 interface,info ether1 link down 10:22:24 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:38 interface,info ether1 link down 10:22:40 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:43 interface,info ether1 link down 10:22:44 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:48 interface,info ether1 link down 10:22:50 interface,info ether1 link up (speed 2.5G, full duplex) 10:22:53 interface,info ether1 link down 10:22:55 interface,info ether1 link up (speed 2.5G, full duplex) 10:23:07 interface,info ether1 link down 10:23:13 interface,info ether1 link up (speed 2.5G, full duplex) 10:23:15 interface,info ether1 link down 10:23:16 interface,info ether1 link up (speed 2.5G, full duplex) 10:26:07 interface,info ether1 link down 10:26:10 interface,info ether1 link up (speed 2.5G, full duplex)
Could you share full script as you got working now if its not a problem ?Same here, I changed toand things worked again.Code: Select all/tool fetch url="..." as-value output=user
:global IPv4addr
:global IPv6addr
:global IPv6prefix
:global DDNSforce
:local WAN4 "1U1-FTTH"
:local POOL6 "1U1-pool6"
# get actual IPv4 address
:local actualIPv4 [/ip address get [find interface="$WAN4"] address]
# check if WAN IPv4 address did change or update is forced
:if ($DDNSforce = "YES" or [:tostr $actualIPv4] != [:tostr $IPv4addr]) do= {
:if ($DDNSforce = "YES") do= {
/log info ("DDNS-check: WAN IPv4 forced DDNS update to $actualIPv4")
} else= {
/log info ("DDNS-check: WAN IPv4 address changed from $IPv4addr to $actualIPv4")
}
# strip netmask from IP address
:local IPv4 [:pick $actualIPv4 0 [:find $actualIPv4 "/"]]
# update Hurricane Electric DDNS IPv4 address
:local FURL "https://dyn.dns.he.net/nic/update\?password=xxx-xxx-xxx"
/tool fetch url="$FURL&hostname=host1.foo.bar&myip=$IPv4" as-value output=user
/tool fetch url="$FURL&hostname=ipv4.host1.foo.bar&myip=$IPv4" as-value output=user
/tool fetch url="$FURL&hostname=host2.foo.bar&myip=$IPv4" as-value output=user
# store new IPv4 address for later comparisment
:set IPv4addr $actualIPv4
}
# get actual IPv6 address
:local actualIPv6 [/ipv6 address get [find comment="$POOL6" interface="LAN"] address]
# check if WAN IPv6 address did change or update is forced
:if ($DDNSforce = "YES" or [:tostr $IPv6addr] != [:tostr $actualIPv6]) do= {
:if ($DDNSforce = "YES") do= {
/log info ("DDNS-check: WAN IPv6 forced DDNS update to $actualIPv6")
} else= {
/log info ("DDNS-check: WAN IPv6 address changed from $IPv6addr to $actualIPv6")
}
# strip netmask from IP address
:local IPv6 [:pick $actualIPv6 0 [:find $actualIPv6 "/"]]
# get IPv6 prefix from IP address
:local pos [:find $IPv6 ":" 0]
:set pos [:find $IPv6 ":" $pos]
:set pos [:find $IPv6 ":" $pos]
:set pos [:find $IPv6 ":" $pos]
:set IPv6prefix [:pick $IPv6 0 $pos]
# update Hurricane Electric DDNS IPv6 address
:local FURL "https://dyn.dns.he.net/nic/update\?password=xxx-xxx-xxx"
/tool fetch url="$FURL&hostname=host1.foo.bar&myip=$IPv6prefix::1" as-value output=user
/tool fetch url="$FURL&hostname=ipv6.host1.foo.bar&myip=$IPv6prefix::1" as-value output=user
/tool fetch url="$FURL&hostname=host2.foo.bar&myip=$IPv6prefix::2" as-value output=user
# store new IPv6 address for later comparisment
:set IPv6addr $actualIPv6
}
# reset forced DDNS update flag
:set $DDNSforce "NO"
{ :local WAN4 "1U1-FTTH" :local POOL6 "1U1-pool6" :global DDNSforce :global IPv4addr :local actualIPv4 [/ip address get [find where interface="$WAN4"] address] :set actualIPv4 [:pick $actualIPv4 0 [:find $actualIPv4 "/" -1]] :if (($DDNSforce = "YES") or ($actualIPv4 != $IPv4addr)) do= { :if ($DDNSforce = "YES") do= { :log info ("DDNS-check: WAN IPv4 forced DDNS update to $actualIPv4") } else={ :log info ("DDNS-check: WAN IPv4 address changed from $IPv4addr to $actualIPv4") } :local FURL "https://dyn.dns.he.net/nic/update\3Fpas ... xx-xxx-xxx" /tool fetch url="$FURL&hostname=host1.foo.bar&myip=$actualIPv4" output=none /tool fetch url="$FURL&hostname=ipv4.host1.foo.bar&myip=$actualIPv4" output=none /tool fetch url="$FURL&hostname=host2.foo.bar&myip=$actualIPv4" output=none :set IPv4addr $actualIPv4 } :global IPv6addr :global IPv6prefix :local actualIPv6 [/ipv6 address get [find where comment="$POOL6" and interface="LAN"] address] :set actualIPv6 [:pick $actualIPv6 0 [:find $actualIPv6 "/" -1]] :if ($DDNSforce = "YES" or [:tostr $IPv6addr] != [:tostr $actualIPv6]) do= { :if ($DDNSforce = "YES") do= { :log info ("DDNS-check: WAN IPv6 forced DDNS update to $actualIPv6") } else={ :log info ("DDNS-check: WAN IPv6 address changed from $IPv6addr to $actualIPv6") } # *** unchecked part start *** :local pos [:find $actualIPv6 ":" -1] :set pos [:find $actualIPv6 ":" $pos] :set pos [:find $actualIPv6 ":" $pos] :set pos [:find $actualIPv6 ":" $pos] :set IPv6prefix [:pick $actualIPv6 0 $pos] # *** unchecked part end *** :local FURL "https://dyn.dns.he.net/nic/update\3Fpas ... xx-xxx-xxx" /tool fetch url="$FURL&hostname=host1.foo.bar&myip=$IPv6prefix::1" output=none /tool fetch url="$FURL&hostname=ipv6.host1.foo.bar&myip=$IPv6prefix::1" output=none /tool fetch url="$FURL&hostname=host2.foo.bar&myip=$IPv6prefix::2" output=none :set IPv6addr $actualIPv6 } :set DDNSforce "NO" }
Does not care. If you read about the problem, it appears above.Before investigate if is roueros problem or not, is better write script correctly...
What did the /tool fetch line have before your change?
/tool fetch keep-result=no mode=https url="$FURL&hostname=host.boo.bar&myip=$IPv4"
(Always care write the script correctly...)Does not care. If you read about the problem, it appears above.Before investigate if is roueros problem or not, is better write script correctly...
Well the difference between is "keep-result=no" vs "output=none". In theory... keep-result=no should have ignored the default output=file & do the same thing... At least in my reading of the docs.What did the /tool fetch line have before your change?Code: Select all/tool fetch keep-result=no mode=https url="$FURL&hostname=host.boo.bar&myip=$IPv4"
Agree. I can downgrade to 7.7 and it works, upgrade back to 7.8b3 and it failes like others here say.Well the difference between is "keep-result=no" vs "output=none". In theory... keep-result=no should have ignored the default output=file & do the same thing... At least in my reading of the docs.Code: Select all/tool fetch keep-result=no mode=https url="$FURL&hostname=host.boo.bar&myip=$IPv4"
Since multiple folks say this used to work, something did change here.
The permission context could be different between scheduler vs script is another thing that's possible.Also it runs multiple time via the scheduler when once run by hand. Then fails again.
We are getting nearer: I changed "as-value output=user" to "output=none" and it failed. E.g. the very first fetch aborted.I think "output=none" is clearer, and theoretically avoids creation of the results array in memory from using "as-value output=user" (to then not use the array created).
Can't as I switched to 7.8b3 because of the DNS/NOERR fix from 7.7.I guess you could confirm if your original code worked in 7.8beta2...
works here ...please check
SUP-107205: dns-static CNAME not working.