Diagnose why we have so many slow nodes on the network

[shaupt131]

No description provided.

[storjrobot]

This issue has been mentioned on Storj Community Forum (official). There might be relevant details there:

https://forum.storj.io/t/discussion-on-scaling-the-network-a-multi-node-sno-tier-proposal/25217/22

[littleskunk]

I originally started to work on this ticket because I had the feeling my storage nodes sometimes count as offline when they shouldn't.
I found 2 issues and created bug reports for them:
#6799
#6786

One small improvement that will help as well:
#6795

But overall my storage nodes are looking healthy. Full story can be found here: #6788

There is maybe one small improvement I can think of. The moment my external IP changes my nodes might be unavailable for about an hour. We could reduce that down to a few minutes. I will talk in my next 1on1 with JT about that.

Up next I am planning to test the performance of each storage node. How many nodes return errors? Maybe that count has increased over time? Is there a set of nodes that is slower than others? I know Marton has some code for this.

Close this ticket and I come up with a plan for the next ticket?

[shaupt131]

Close this ticket and I come up with a plan for the next ticket?

Yes please. Also, do the 3 tickets you created above belong to the satellite team?