Cloud DNS is unavailable?

As I can see over my company network, a lot of routers can not register itself in you Cloud DNS.

As I try to traceroute to cloud.mikrotik.com, the trace won’t go to the host, and this is at least from two different directions (that is, my routers are at al least two different ISP, these ISPs are using different uplinks so the directions to your hosts are different, too).

Here is the first trace (I drop “my” hosts from the listing):

1 - skip -
2 109.239.209.41 0% 1 1.1ms 1.1 1.1 1.1
3 188.254.100.133 0% 1 0.9ms 0.9 0.9 0.9
4 95.167.93.150 0% 1 35.6ms 35.6 35.6 35.6
5 92.50.199.30 0% 1 46.7ms 46.7 46.7 46.7
6 195.122.0.213 0% 1 45.3ms 45.3 45.3 45.3
7 100% 1 timeout
8 100% 1 timeout

Here is the second trace:

1 83.219.128.0 0% 1 1.5ms 1.5 1.5 1.5
2 10.192.168.145 0% 1 1.5ms 1.5 1.5 1.5
3 10.192.168.255 0% 1 8.3ms 8.3 8.3 8.3
4 91.90.249.130 0% 1 8.3ms 8.3 8.3 8.3
5 195.246.227.8 0% 1 8.3ms 8.3 8.3 8.3
6 89.201.51.214 0% 1 8.8ms 8.8 8.8 8.8
7 100% 1 timeout
8 100% 1 timeout

Here is another trace:

1 - skip -
2 79.104.235.78 0% 2 58ms 58 58 58
3 213.221.4.132 0% 2 57.5ms 58.6 57.5 59.7
4 195.122.0.213 0% 2 69.1ms 68.5 67.9 69.1
5 100% 2 timeout
6 100% 1 timeout

Another trace:

1 - skip -
2 217.16.23.138 0% 1 97.2ms 97.2 97.2 97.2 0
3 193.232.245.156 0% 1 22.2ms 22.2 22.2 22.2 0
4 195.122.0.213 0% 1 23.1ms 23.1 23.1 23.1 0
5 100% 1 timeout
6 100% 1 timeout

195.122.0.213 looks like the most often ‘last’ hope.

The problem is there for about 30 hours so far. Is there any chance to see the cloud system running soon, and should we rely on cloud feature in the future?

P.S. Just in case, do you have any availability checks of your “cloud” system from around the world?

Same here in Hungary

Yes, get a servfail:

; <<>> DiG 9.9.5-9+deb8u4-Debian <<>> a xyz.sn.mynetname.net.
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: SERVFAIL, id: 51194
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 512
;; QUESTION SECTION:
;xyz.sn.mynetname.net.	IN	A

;; Query time: 2017 msec
;; SERVER: 8.8.8.8#53(8.8.8.8)
;; WHEN: Sun Jan 03 12:55:50 CET 2016
;; MSG SIZE  rcvd: 58

Also see last couple of posts in http://forum.mikrotik.com/t/problem-with-mount-point/94/1 for history.

I’m posting from South Africa and the service is also not working from South Africa.

If I use a tool such as OpenDNS cache check it seems like it is working from some countries.

https://cachecheck.opendns.com/

Working from:
Chicago, Illinois, United States

All other OpenDNS servers also report “SERVFAIL”.

No, I can’t even connect to that host at all. Pretty weird situation, a lot of hosts are rely on their dyndns (cloud) service and it won’t work at all.

I think we need to develop our own DNS service to overcome such crazy things. Mikrotik offered Cloud so many people used to use it as part of their infrastructure but it is not that always-on as should be :frowning:

This is the first time since starting to use the MikroTik Cloud “DDNS” service that I am experiencing an outage.

I have also e-mailed support[at]mikrotik.com - no answer.

I also sent them a Tweet now to @mikrotik_com

Just wish someone from MikroTik would get involved here and give us some for of feedback or status update. I tried finding a status page on their website. Anyone know if MikroTik has a status page for this type of thing?

I’m afraid no, the don’t, and I also think they decided to prove that low-end routers and also comes with ‘nothing to trust’ services. Very sad.

It’ll be much harder to recommend their solutions to people.

I wrote it occasionally several times here on the forum. Mikrotik ddns service doesn’t offer nothing special, actually is too simple and not configurable. The only advantage is that it doesn’t need any script to run it in ros. All other things, reliability included, are cons. I have been with dnsexit for years without any outage and with full features of dns.

Yeah, you’re right, but with mass deployment this out-of-box ddns is quite handy, and, moreover, it won’t need anything but the single tick in the GUI.

Thanks Mikrotik they have scripting added as well.

cloud.mikrotik.com resolves itself to their office IPs (two ISP = 2 different IP), and I bet they have some Mikrotik router as a border one. Maybe they ruined the config just before the holidays (or maybe router reseted its config itself - once I’ve seen such an unbelievable behaviour), so someone have to go to the office to fix it and noone care.

Shame on them. It worth thousand advertisement words to show once such an outage with no traces that anyone tries to fix it after a day of outage. Big impact on sales, I guess.

No one at Mikrotik listening? This is the 3rd time we have a complete outage of this DDNS service.

Once I can reach my system again there is nothing but migrating away. I wrote to support at mikrotik.com. Shouldn’t such an infrastructure be monitored?

This is no getting critical for us as the business week starts tomorrow..

Rémy

http://forum.mikrotik.com/t/did-we-loose-ip-cloud/81616/132

Service is back online.

How often do these cloud servers go down guys?

Nit that often, but the problem was it was down for days, no reaction from support via mail or on forum, and even no excuse after they arrived to the office and rebooted that damn server.

Yes, this is cheap “under my desk” machine acts as server and installed in their office. No real cloud infrastructure :open_mouth:

So it is wise to set up another dyndns via script together with MT cloud things.