Full node, no egress

I’ve set STORAGE=500GB on this node when it was at 750GB to stop it from growing and since then (4 days) there has been no egress. I do see a few “piecestore deleted” in the logs.

I don’t wanna say that it’s broken but how big are the odds? I have another full node on the same machine which has >0,5% of its storage in daily egress

Welcome
Let me get this straight you started with allowing for 750gigs then after it filled you just set it to 500gigs?
Also when you have more then 1 node on the same isp it acts as one node which means you wont get different data for both they pretty much share anything coming in.
Are they both new nodes because this screen shot showing you just started the node this month, You can’t expect there to be egress within the first month of running your node.

Let me get this straight you started with allowing for 750gigs then after it filled you just set it to 500gigs?

I set it to 3.5TB then I decided I preferred this disk for something else I set it to 500GB which is the lower limit.

Also when you have more then 1 node on the same isp it acts as one node which means you wont get different data for both they pretty much share anything coming in.

That’s clear, I have 3 non-empty nodes which are almost identical usage-wise.

Are they both new nodes because this screen shot showing you just started the node this month, You can’t expect there to be egress within the first month of running your node.

That makes no sense.

  1. I had the usual 0.5% egress on this node until I reduced the storage
  2. Data is not whiskey, it does not get better with age. Data is used mostly soon after it’s stored and increasingly less as it ages.

I see 2 logical explanations:

  1. There was egress before because some users are uploading then immediately downloading, like a sharing service or checking if it went well. Maybe I can view this in the logs.
  2. Something in the code which affects egress of a full node, either intentional or unintentional.

I’m not complaining, just sharing my observations.

Sorry but this is false the data your node got is testing data and its not predictable of how the data gets redownloaded. Which you would see which satellite the egress came from which my guess is saltlake which is a test satetllite. The more storage your node has the more chances that it will get redownloaded at some point, Data stored on a new node does not get redownloaded right away.
If your node sees customer data this is 100% unpredictable so you won’t know when they will access it.

Sorry but this is false the data your node got is testing data

I assume at this point there is also production data? There have been Tardigrade clients for some time now. Does their data only get stored on vetted nodes?

and its not predictable of how the data gets redownloaded.

Define “predictable”. Theoretically all the neutrons in a nuclear reactor could miss and the reactor will shut down.

I find this pretty predictable (other full node)

I hope you can see that your explanation doesn’t seem complete.

Your obviously not showing the same node. You posted about your new node.

One node is older the new one is just getting test data. Your new node is still being vetted and of course vetting nodes arent going to see the same amount of customer data as say a fully vetted node(trusted)

Like I said before Its 100% not predictable Your full node is older and has more data, Your new node is vetting is getting pretty much testing data from saltlake satellite.

Id like you to define “predictable”

1 Like

In the mean time I did get a few download requests hahah

2020-04-22T18:28:04.586Z INFO piecestore downloaded {“Piece ID”: “24EDWVXHPZBHYLFREATN5OKCDHP5EMHBWXZ64D4BQKP7XNWKCQGA”, “Satellite ID”: “12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S”, “Action”: “GET”}