It looks like space is still getting filled at the long-term-normal rates we’ve seen in the past (around 8TB-per-/24-per-year?). But it can be hard to notice beneath the mountain of expiring test data.
Ignore Storj unless you get an email saying a node is offline: it’s like watching water boil: do something else with your life while you wait 
I understand what you are saying, but in my situation I need to monitor performance and provide both internet and power backup.
This naturally consumes resources…
I am forced to monitor and asked a question to understand if there are plans to add more test data or increase the load in the near future. (I am not whining or crying, I am asking for objective data and plans if such are known)
Yes, I missed that, thanks.
Then we have time to optimize our nodes and hope for good fill rates to resume (because we have nothing but hope).
From the request to publish the graph here and include test data for 1 day per week (publish the graph in advance) so that those who are doing optimization can test their nodes and results under load.
Thanks, have a good mood everyone.
Yes, I’m thinking about that too. I just do not know, how it could be implemented. For example, the most useful thing is to show a success score (from the satellites) somewhere, but it’s dynamic, also node- and customer- dependent, how to collect, how to show?? I have no idea.
Perhaps it’s almost the same as a local Success rate script - Now updated for new delete terminology since v1.29.3, but I do not know for sure…
They did some tests by boiling Cesium or some metal; when someone was looking at the melted stuff, the boiling point was reached later than when nobody was looking.
So… if you don’t look, things are happening faster. 
Yeah, right.
And they just did it by boiling it first watching. Let it cool down, and did it again without watching.
It was statistical significant of course…
Reference or it didn’t happen…
It looks like tests returned, but now on US1 and EU1 
I’m not aware of such tests from Storj side, so I would assume that this is a natural customers’ load.
Things looking pretty normal as of this morning.
I do have many terabytes of recently garbage collected hard drive space ready to accomodate a scaled up test though, lol.
I saw increased traffic for a couple of hours, but seem to be back to baseline now.
Might have been some client-side testing, perhaps?
Back to waiting patiently…
I have some increase on 40%, but it is not like with tests, where almost constant load, it has peaks and drops, so I guess a normal clients traffic.
It has been some wild months! Some weeks had over 25TB ingress for me. Many had over 5.
My scripts does not account for non-yet-marked-for-trash trash, aaaaaaaaaad then there was the v1.111.4 that corrected taht issue. It looks like we’re just about back to normal operation - and that’s great
All,
I wanted to post a thank you and an update. Thank you all for the wonderful discussion that has been happening in the thread. We are excited about the growth we are seeing, and appreciate all the help you have given us to prepare for it. As has been indicated our pipeline of customers continues to grow and some of our larger deals are progressing and we anticipate material growth in customer data soon. Unfortunately enterprise sales process sometimes move slower than we’d like them to and so there end up being un anticipated delays. We are experiencing a bit of that now which is why we stopped uploading test data to SLC to create space for customer onboarding, but have not yet onboarded the new customer. I wish I could be more definitive regarding when exactly things will proceed, but unfortunately all I know is that it could be any day. Thanks again for all your support, we are excited to grow with you.
Bryan
The delayes are welcomed to give us time to fine tune our storagenodes, to catch some bugs and update some hardware. These tests were verry helpful and I’m pretty sure they made our network more reliable and more performant. It’s better to embarq big clients one by one and to evolve step by step. It allows improvements to be made by Storj team and SNOs.
10PB from one customer, holy moly! 
thats like 690 disks of 16TB nominally, but it must be just a first toe tip to check the water, for sure they have much much more, if they like us!
So get your nodes together people, who haven’t done it yet.
Thank you for the update.
To be honest, I kind of miss having to debug and fix the random things that would break when we had all the test traffic.
in the medium-long term, it might be nice to have a test satellite (even, dare I say, unpaid) that an operator can easily opt in or out of just to practice bombarding their node with traffic.
The opt in/opt out/ opt in for the test sat whould be a nice feature, maybe with a delay like a month or 2, to prevent abuse.
You can do it right now with new identities, but it’s too much hassle.
