[Solved] "License key usage exceeded"

Unimus support forum
Post Reply
SlothCroissant
Posts: 2
Joined: Wed Nov 17, 2021 2:16 am

Wed Nov 17, 2021 2:26 am

Hey there,

I purchased a few licenses today for Unimus, and got some routers/switches setup and working. All was super smooth, and I'm really impressed. However, several hours later (at around 00:00 UTC), Unimus just.... stopped working. When I try to login to my local deployment, it tells me that my license key usage has been exceeded, and I need to purchase more or enter a new license key.

Furthermore (and honestly more worrisome to me) is that the logs on the app itself (deployed into Docker) state that it has disabled itself entirely due to this licensing issue - and is no longer protecting/backing up my configurations. I tried to login to the portal to check my licenses are still there, but that's throwing a 500 error.

Code: Select all

unimus_app.1.djf0ers1ditf@docker02.lan    | 2021-11-16 15:58:09.004  INFO 12 --- [nio-8085-exec-4] net.unimus.business.aaa.AAAHandlerImpl   : User 'user' logged out.
unimus_app.1.djf0ers1ditf@docker02.lan    | 2021-11-16 18:01:31.917  INFO 12 --- [        qrtz--3] net.unimus.system.UnimusImpl             : Setting new license key state 'AMOUNT_EXCEEDED'
unimus_app.1.djf0ers1ditf@docker02.lan    | 2021-11-16 18:01:32.418  INFO 12 --- [pool-2-thread-1] net.unimus.business.scheduler.Scheduler  : Un-scheduling 'discovery / backup' job 'At 3:00 AM, every day'
So my ask is this:
  • Why did Unimus just randomly decide I was over my license limits, 3+ hours after I first set everything up and things were working great?
  • If the portal being down coincides with my licensing issue, why is it that the app will stop working when there are server-side issues? Shouldn't it cache license info for some time, etc to survive issues such as this?
  • How does one log a support ticket if portal.unimus.com is down?
Hoping there are easy answers to these, as the application itself looks super useful (coming from RANCID & Oxidized, myself), and I'm excited to potentially expand usage of it.

Thanks!
User avatar
Tomas
Posts: 1206
Joined: Sat Jun 25, 2016 12:33 pm

Wed Nov 17, 2021 2:41 am

Greetings,

Apologies, this is an issue on our end caused by an external system. Our payment gateway was experiencing issues tonight and returned wrong transaction results to our Portal. This caused a few transactions to fail validation, and the related subscriptions to be rolled back.

We are fixing this right now (and our Portal is down due to this). Your local Unimus instance will not be affected in any way, and will be back to functional as soon as we resolve this issue.

Only customers who purchased a license in the last 24 hours which still had an outstanding transaction validation were affected (you being one of those). This normally has no chance of happening, and you had the unfortunate "luck" to be one of the very few (0.004% of user-base) affected by this.

Apologies again.
SlothCroissant
Posts: 2
Joined: Wed Nov 17, 2021 2:16 am

Wed Nov 17, 2021 2:46 am

Hey there Tomas,

Really appreciate the quick response and clarification on the details. Will wait for things to clear up and hopefully can continue testing tomorrow.

Have a great evening!
User avatar
Tomas
Posts: 1206
Joined: Sat Jun 25, 2016 12:33 pm

Wed Nov 17, 2021 3:45 am

A small update / post-mortem for anyone interested:

Our Portal is now back up, and we are now seeing correct responses from our Payment Gateway. This was a global issue at the Payment Gateway we use, where (as I mentioned in the previous post) it returned wrong results for some transaction statuses.

For a little bit of background - purchase transaction (on any online platform) don't immediately "succeed". Banks initially "approve" a transaction, but the transaction is not fully "cleared" after this approval. It gets an initial approval and then moves to a "settling" state. Our Portal assigns licenses to accounts immediately after a transaction is approved (so you don't have to wait for your license to become active), and performs checks every hour until a transaction fully clears. In cases where a transaction fails to settle (which can happen), a Subscription is rolled back.

What happened tonight is that due to the Payment Gateway issues, transactions which were not fully cleared yet - specifically a few (not all) of transactions that happened in the last 24 hours, reported settlement failure and our Portal rolled back Subscriptions for a few customers.

How this affected local Unimus instances:

Normally, local Unimus instances cache their licenses, so even in cases of outages of our licensing infrastructure, our customer's (your) Unimus instances would not be affected. However, there was no outage on our end, rather Subscriptions were "wrongly" rolled back (due to the above described Payment Gateway issue). Local Unimus licenses synced the "invalid" license state and locked themselves. Normally this only happens when someone cancels their subscription, in which case their Unimus instance will stop working as expected.

When this happens, your Unimus instances do NOT in any way delete or touch any local data. So when the local instance gets an "all OK" back from the licensing server, it will resume it's working in exactly the same state and with the same data as before it locked itself. So good news is that there are no repercussions (other than temporary Unimus inaccessibility which we apologize for) for the affected customers.

We identified the exact accounts this affected - luckily only 0.004% of our user-base was affected. We really do apologize that this happened. As is common with issues like this, we never imagined a failure scenario like this could occur. We will take steps to mitigate issues like this into the future (will implement more checks and validation logic to make sure a Payment Gateway issue can not affect customers in this fashion).

We are still fixing the rolled-back accounts, and should have them back to working order soon.
User avatar
Tomas
Posts: 1206
Joined: Sat Jun 25, 2016 12:33 pm

Wed Nov 17, 2021 5:47 am

We have just finished fixing the affected account - everything should be back to fully functional again.
Post Reply