The Global Downfall of Ubiquiti UniFi
Yesterday, August 31, around 5:30 p.m., users began to complain about the operation of the cloud system of their devices, and that is that they could not connect remotely to all devices using the UniFi platform in the cloud. At 6.30 p.m. the manufacturer Ubiquiti confirmed that users began to report problems, and began to investigate the problems. Of course, with this crash, neither the Cloud access from the web nor through the UniFi mobile applications worked, they could not even access the main web.
Two hours later, around 8:30 p.m. they located the problem, it seems that there was an availability problem with the AWS service (Amazon Web Services) and this could be the cause of the connectivity problems, of course, the Ubiquiti team began to monitor the situation to investigate if it was an AWS problem or your own. Around 10:30 p.m., the Ubiquiti team confirmed that the service had been partially restored, although it is possible that some users could still experience connection problems, and that they would try again after a while. Officially, at 11:00 p.m. Ubiquiti confirmed that the service had been successfully restored and there should be no problem.
Thanks for your patience. A current AWS outage could be impacting https://t.co/p8qZSOzCtF services. Some users may not be able to connect remotely to devices via https://t.co/5E8K0cDfM2 or the UniFi mobile apps. We will provide periodic updates at https://t.co/U9bx3dy5kj
– Ubiquiti Inc (@Ubiquiti) August 31, 2021
Users were between 4 and 6 hours without service in the Ubiquiti cloud, therefore, they could not access remotely to manage, administer, nor could they receive monitoring alerts from the different routers, switches and professional WiFi access points they had. deployed in the cloud, because everything was down. This is one of the main problems of the Cloud, it is true that we have a really high uptime, higher than 99.9%, but we will always have several hours and even a couple of days during a whole year without being able to connect correctly to our networks to manage them, in addition, this has been in the middle of Tuesday and not on the weekend.
The main problem of Cloud management: availability
In RedesZone we have written a complete article explaining the strengths and weaknesses of network management from the cloud, one of the weaknesses to which we referred is precisely what happened here, the availability of the service, and that is that we depend 100% on the good work of the manufacturer, its development team and the engineers to solve the problems . While these problems are solved, we will not have the possibility to deploy new configurations, to modify certain network configurations, we will not even be able to see the monitoring of each and every one of the devices.
Although a fall of this type is not “normal”, it is also true that we will always have certain time intervals where we will have a stop for maintenance of the Cloud, update of certain parameters in the cloud, or falls due to AWS or wherever they are hosting the servers in the cloud. It is possible that these “scheduled shutdowns” just harm you because you are going to install new devices, or manage those already installed, but it is one of the weak points of the Cloud: management and availability is not in your hand.
Finally, another very important aspect is the network management security from the cloud, and is that we can only authenticate with our credentials and a second authentication factor, no service such as VPN is used in order to facilitate management, and of course, we delegate to the manufacturer all the security regarding DoS attacks and authentication.