V7.22rc [testing] is released!

I see what your saying. I see the certificate.

echo | openssl s_client -connect acme-v02.api.letsencrypt.org:443
Connecting to 172.65.32.248
CONNECTED(00000003)
depth=2 C=US, O=Internet Security Research Group, CN=ISRG Root X1
verify return:1
depth=1 C=US, O=Let's Encrypt, CN=E8
verify return:1
depth=0 CN=acme-v02.api.letsencrypt.org
verify return:1
---
Certificate chain
 0 s:CN=acme-v02.api.letsencrypt.org
   i:C=US, O=Let's Encrypt, CN=E8
   a:PKEY: EC, (prime256v1); sigalg: ecdsa-with-SHA384
   v:NotBefore: Jan 24 20:58:09 2026 GMT; NotAfter: Apr 24 20:58:08 2026 GMT
 1 s:C=US, O=Let's Encrypt, CN=E8
   i:C=US, O=Internet Security Research Group, CN=ISRG Root X1
   a:PKEY: EC, (secp384r1); sigalg: sha256WithRSAEncryption
   v:NotBefore: Mar 13 00:00:00 2024 GMT; NotAfter: Mar 12 23:59:59 2027 GMT
---
Server certificate
-----BEGIN CERTIFICATE-----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-----END CERTIFICATE-----
subject=CN=acme-v02.api.letsencrypt.org
issuer=C=US, O=Let's Encrypt, CN=E8
---
No client certificate CA names sent
Peer signing digest: SHA256
Peer signature type: ecdsa_secp256r1_sha256
Peer Temp Key: X25519, 253 bits

Already added that missing certificate months ago. I also posted about it.

Thanks

@CGGXANNX would you know how to create such a certificata by chance or if there’s a write up on the net.

How it get a Let's Encrypt certificate? That should be well documented...

I was referring to a the Server certificate in the certificate chain.

Anyone experiencing huge Cake CPU loads compared to previous versions?

I wonder what might have caused this, since there’s nothing in the changelog related to BFD.

BFD has been broken before without related topic in the changelog. Either the BFD maintainer does not keep changelog entries very well, or it gets broken by totally unrelated changes.

What is still worrying is that there apparently is no regression test suite. @mrz claimed there was no problem and he could make BGP connects without issue, but apparently the setup where he tests that has no BFD. Seems easy to enable that at least on a link in a test network used to validate new versions before release…

When @kleshki had not pointed this out, 7.22 would probably have gone into release with this serious and dangerous breakage…. well, it can still happen of course.

When using a "normal" ACME client (e.g. certbot on linux), it pulls 3 files when getting new/refreshed certificate: private key, certificate file and "full chain" certificate file. One should use key file and full chain certificate file in server environment, not the "simple" certificate file.

I see from the updated documentation that the default rules will be gone, and the order would be specified by /routing settings set policy-rules=. This is much better and solves the export/import problem.

It seems that user defined rules can be inserted at only one position in the list. Which probably means that users can still create routing rules with the new actions and flags (that's why WinBox compatibility was updated in rc2), so that when they want custom routing rules inserted at different locations (two lookup rules surrounding mangle for example), they would remove some of the pre-defined items from policy-rules, and insert them back as rules under /routing rule?

That seems to be a bad solution even compared to what we have now, as it defeats flexibility previously given. It is probably better to go on with what is now and accept the breaking change, rather than stop a good feature to be released

Probably what currently available in rc2 can easily be replicated by setting policy-rules=user (and nothing else), then manually adding back the beta/rc1/2's default rules to /routing rule in whatever order you wish? I don't think there is any loss of flexibility.

EDIT: of course, the rules must be populated first before switching to policy-rules=user, otherwise IP connection to the router will be loss :smiley:.

Ah, if that works that way then probably it is good, I’ve misunderstood the concept

Lol, I hope upgrading from any 7.22 version is going to work flawlessly.

It is not limited jut to one “user“ chain, you can add any amount of chains and insert anywhere in the policy-rules list.

Interesting, but if I do:

/routing settings set policy-rules=user,vrf-lookup,vrf-unreach,local,user,mangle,main

how can the router know which rules are for the 1st user occurrence and which are for the 2nd one? Or we can give custom name for the chains?

I guess we'll know more with rc3.

you can give custom name

How is everyone’s wiregaurd speeds after upstream updates? I noticed that speeds droped for me in half but only in one direction, so im getting 1000/500, it was 1000/1000 before just fine.

I like to new approach to exposing the "builtin route rules".

I notice the docs also were updated to use "Marvell Presta" instead of long (sometime partial) list of models. This seems like a good change for doc readability. But tiny suggestion on this... is the table of Marvell Presta features include whether it has PTPv2 support, since not all models (AFAIK) support PTPv2. Similar with QoS on switches. There still too many tables to cross-reference on the switch chip differences. See Marvell Prestera switch chip features - RouterOS - MikroTik Documentation with link to other pages:

From my point of view, this custom namd is where the flowspec rules and dynamic pbrs will be fitted.

If that's the case. That seems like a good thing.

@mrz any insight on IPsec VTI if this would see the light of the day in this release or future release and also do you guys consider sdwan in the future most of the tech is already available in ROS afaics

It is in rc status already so for sure not in this release, such a major feature would not be added while in rc status.

So again, we have to hope for the next release…