Almost no traffic at the end of the month?

Since about 8 AM UTC, 31 Jan, my nodes receive barely any traffic.

At was very similar between December and January.

Why is this happening?

root@storj01:~# vnstat -h
 eth0                                                                     14:06 
  ^   t                                                                         
  |   t                 t     t              t     t                            
  |   t              t  t  t  t  t  t     t  t  t  t  t                         
  |   t  t     t     t  t  t  t  t  t     t  t  t  t  t                         
  |   t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t                         
  |   t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t                         
  |   t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t                         
  |   t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t                         
  |   t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t                      
  |   t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t  t                      
 -+---------------------------------------------------------------------------> 
  |  15 16 17 18 19 20 21 22 23 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14

The mystery of customer usage patterns!

2 Likes

Is there any customer traffic yet? Or is it mostly testing by Storj labs themselves?

It’s both. For now still a lot of Storj, but that’s likely going to shift over the upcoming months.

Yesterday a bug occured. Nodes were using more storage than was allowed by SNOs.
Devs fixed it, but turned off test traffic for the moment.

1 Like

Hello i have no traffic also
I would like to stop my node because at this time i will loose my money…

I run a cluster of tower PC for one node and this consume very very very important capacity of electric energy.

only 300-400GB a month once my node was vetted…

@Storj what can you answer to this problem and to the fact i am not profitable ?

However my node is full performance ready… SATA 3GB with a 300Mb fiber uplink… MTBF will ruin my profit…

We are holding off on downloads since it it nearing the end of the month and we are planning to set up a new satellite for stress testing
Low traffic for few hours, stefan-benten satellite data delete?

1 Like

However, the future expectation is that customers will be delivering most of the load, and how much you get will be dependent upon their usage patterns

It has always been said, don’t buy hardware just to run a node and run a node only on hardware which is running anyway. So electricity wouldn’t be a factor.

That being said, with all the testing going on recently and the surge payment on top I can’t see how anyone wouldn’t make profit at the moment or at least cover the electricity.

4 Likes

I’m sorry but I’m an SNO with 5 nodes, being the oldest one a RPI 3B+ on a Storj Signature box, running since May 2019 in The Netherlands.
I believe something major, changed as my “upload failed” rate went to the roof.
I upgraded my ISP to 100 download-40 upload very recently, have available disk space as well.
My “upload” success rate was 96%. It dropped to 13%!
So, definitely, something structural changed, don’t know what, but did see this behaviour on my other 4 nodes.
And I’m not referring to test data or real data…I’m just talking DATA pure and simple and only started after 0.30.12! Coincidence?
Don’t know, but I’ll keep monitoring.

In order to provide the highest speed possible, the amount of traffic you receive can depend on the geographic location of users. As we launch into production and begin to grow, hopefully we can bring on users in many different parts of the world.

Yes, and that makes totally sense. My issue is that from one release to another the whole upload pattern changed and my upload success rate went from 96% to 10ish solely based on location and internet speed?
Sorry but on a structural perspective, either the targets changed or something big changed in the code to “filter” out a few SNO based on hardware performance or some very big player just came in the market…just my 5 cents of course!

If your node is more than 6 month old you can call graceful exit and get back your holdback amount. Alternative you can just shutdown your storage node and let the system handle it via repair. You would lose the holdback amount. In both cases it is a one way ticket. You can’t cancel the process.

My plan is to wait a few months. With surge pricing I will get a lot of money for January. That is more than enough money to go through a longer time periode with low traffic.

2 Likes