I have usage collection disabled in the GUI settings:
But TrueNAS still seems to be phoning home every ~30 minutes for a couple of times, and then retries after ~3-4 hours:
How do I make it stop?
I’ve just replicated the same after disabling usage statistics. It may or may not be a bug, someone may know more about how it works under the hood. I see @kris has commented on a similar post on Reddit.
If you’re super concerned, I’d just add 0.0.0.0 usage.truenas.com to the hostname database in Network → Global Configuration
Personally I wouldn’t be too concerned about the usage statistics, I keep them on as it’s incredibly useful for development of the product to know what features are most used, or what does not see any usage at all.
1 Like
I wouldn’t mind sending statistics, if I could audit them first. Then I would simply allow the domain through Squid, indirectly solving the issue.
But since I already have usage statistics disabled, it’s just annoying that Squid logs are being flooded with bogus requests.
I’m not sure if 0.0.0.0 usage.truenas.com would prevent TrueNAS from trying the proxy first, so I’d likely still be left with bogus requests in the logs.
It is absolutely NOT that I distrust ixSystems, however I am paranoid about my privacy in this web 2 age where my personal data is essentially Google/Facebook/X profits, so I have Usage Collecton disabled.
So it is very disappointing to learn that usage details are still being sent to them despite having unchecked this box. This sort of thing is pretty damaging to users trust in ixSystems.
In addition, like others who have commented here I would be more likely to voluntarily donate my Usage Statistics to ixSystems if I had a better idea of what is going to be shared. Here is the tooltip for the Usage Collection checkbox:
This is not exactly a reassurance that personal data stored on my NAS won’t be sent to ixSystems. Yes - it says that the data won’t be tagged with an identifier, but it says absolutely nothing about what data is sent to ixSystems. I more informative tooltip which says what data is sent would be more likely to result in a ticked checkbox.
As hypothetical examples (and I stress that I have deliberately made up these examples to test what is / isn’t sent to ixSystems):
- What if my (private) server name is obnoxious or abhorrent or blasphemous? Is it shared with ixSystems?
- Ditto for pool names? What if my pool name is “[mod edited for decency]” even though it stores only the local weather forecasting data I collect?
- Ditto for dataset names? Or SMB share names?
- Are file names shared?
- Are my user details shared?
- What about details of the Apps I run? Or their process names? (Some people might think this is useful and innocuous data and be happy for it to be shared. Some people might prefer details of their apps NOT to be shared.)
I am also entirely unclear what legal basis underpins this data collection. Here in the EU you need Data Protection consent to collect any data in which you have to state what data you collect and where / how to hold it and process it, and who you will share it with - and this consent can be withdrawn. I can’t remember whether I consented to data collection when I installed TN SCALE or what the exact details about usage where detailed in it, but I cannot see any means of withdrawing such consent or to verify what I consented to.
IMO transparency is key here.
1 Like
TrueNAS has a data collection statement that should provide detail on what is collected. It states while disabled, system “capacity” and software version is still collected, which would explain the calls out to usage.truenas.com (though I have no idea what they mean by capacity).
Though as you mention, transparency is absolutely key, and I’m not seeing a lot of information in this documentation as to what is ACTUALLY collected. Like what exactly is “Usage and Performance statistics”, or “Pool configuration” and what is the scope of collection under these categories?
Hopefully a response from iX soon as it is rather concerning, and it might be a good idea for them to get this documentation updated with an in-depth explaination of what is being collected for those more conscious about their privacy.
1 Like
Actually, I found the Data Collection statement rather comforting, particularly its absolute statement about what is NOT collected.
Here is what I would personally like to see as minor enhancements:
-
The tooltip being enhanced to state that the system version and capacity is always collected but it can be blocked at the firewall without adverse consequences.
-
The tooltip being enhanced with the URL of the Data Collection Statement.
-
The Data Collection Statement table having a third column describing what is always collected.
-
The Data Collection Statement having a new section which gives precise details of the Optional data collected i.e. does pool information contain the pool name which might contain potentially embarrassing or private information.
-
The Data Collection Statement giving detailed instructions on how to block the version and capacity logging.
-
The GUI Settings menu at the top having an additional entry for Data Collection Statement.
-
If it doesn’t already have it, the installer having a popup explanation of the data collection with a checkbox the user needs to check to confirm that they have read and understood the Data Collection summary.
2 Likes
Hey folks,
You can see exactly what’s transmitted to iXsystems by running midclt call usage.gather from an SSH session/shell. I’ll see about working more detail in the documentation, as well as moving to have the tooltip expanded and/or a “More Details” link available for those interested.
We’re not collecting the names of your server, pool, dataset, share, file, Apps, processes - anything like that. The only “Names” we’re collecting are the names of the Apps from the catalog (eg: you’ll see “minio/2.0.6” but not the user-assigned one) - but bear in mind this information would be included in a debug file, such as those requested for a bug report, feature request, or other troubleshooting. So certain dataset names such as that example are probably, uh, “rather ill-advised.” 
Regarding the “anonymous statistics” - we collect your version of TrueNAS, and the total pool capacity of your system (eg: “TrueNAS SCALE 24.04.0, 55 TB”) - nothing more.
3 Likes
@HoneyBadger Just as before, I find your answer somewhat comforting (and it was pretty much what I was expecting).
In reality, my pool names are actually very boring - hdd-pool, apps-pool. And the dataset names are equally innocuous. Indeed, unless anyone wants to find out which cartoons my kids are interested in, then even the file names aren’t going to be of much interest.
But in the interests of transparency and comfort factor, I think that the details of your answer should be part of the UI / help page as suggested.
3 Likes
Checking, I did not find an existing Jira ticket on this subject, so I created one;
https://ixsystems.atlassian.net/jira/software/c/projects/NAS/issues/NAS-128857?jql=project%20%3D%20"NAS"%20ORDER%20BY%20created%20DESC
If you agree with the subject, please vote for it, using the blue thumbs up in the upper right corner. (Whence you click on the blue thumb, it brings up a popup where you can vote for it.)
2 Likes