300 fault files in /storage/blobs folder

Current offers under $15: Intel DC S3500 120GB for sale | eBay

Obviously, individual sellers need to be looked at; but fakes are very unlikely in this segment.

my past purchases



2 Likes

You sound very reasonable I don’t have counter arguments.

Have installed cache ssd - will see how it performs now.

1 Like

This is 50 USD with shipping and import taxes. My local offers are a bit better, but still not even close to 10 USD.

I don’t know where you live.. but aren’t there local electronics recyclers, not necessary with online presence? How do datacenters and corporations get rid of old gear?

1 Like

The local offers are exactly that, used data center hardware. Poland. The cheapest I can find is a 160GB drive for 22 USD with shipping. Though, thank you for pointing this out, still cheaper than I expected.

1 Like

Sure… in a perfect world.

1/10 of 1 cent,
Julio

In a real world too, see

“Neighbors” here - the nearest /24 subnets, not necessarily your real neighbors :smile:

1 Like

You misinterpret the phrase ‘Over time.’ Yes instantaneous checks upon demand as noted by Alexey do occur, and by that virtue do cumulatively help. But the satellite has no ‘memory’ at any time of your success/efficiency or failures of the past. I was pointing out my perception of an over statement, as not to mislead SNOs.

1/50th of a cent,
Julio

1 Like

So, if understood that post correctly - it’s relatively easy to recover from “bad node” as just improving its performance?

I noted this grafana dashboard is useful to track success rate. Setting Up a Dashboard for Commercial Storage Nodes - Storj Docs

1 Like

In my experience it’s on the scale of minutes, not seconds, and of course I assume storj can tweak ot anytime.

Perhaps, but not always. Too many factors, the main of them - the customers’ activity and the node concentration in the customer’s location.