I keep an eye on my monitoring script. All Ok with this nodes. Also where is slow 100mbit connection.
And this is not a router. This is switch.
Router is a top case with 2x xeons 2011
I suggest move on rigs thread.
I have 2 fiber networks from diffternt providest, so I have different subnets it duble trafic. Also when your trafic is separated in lot of hdds i have better responce time and more chanses to win tha race.
Same things with Databases, thay are on SSD or NVMe, so HDDs are dealing only with read/write data what once more give faster writing speed.
1 Like
different subnets it duble traffic.
Nice! that explains 2x more traffic
Same things with Databases, thay are on SSD or NVMe
how did you get databases separated from storage?
Because my storage is created in subfolder of DBs
is there a parameter that you can specify for DBs?
also location can mater.
1 Like
I use windows GUI, and from some version it can be separeted. I thing it is from 1.6.3 verson only.
option in config
Storage2.Database-Dir: C:\Program Files\Storj5
1 Like
Colleagues, please keep thread clean, it not a chat, thank for understanding.
1 Like
his egress was very high… now it’s just returning to the avg…
think about it… over 1% egress … that means 30%+ of the stored data would be downloaded in a month…
sure this is not to regular consumer data usage… but still how much of your data on your computer do you actually use in a month… 30% would be very high… if we can get a consistent 10% egress of stored pr month then we should be thrilled…
I use my toolbox for statistics, i think need to add some % dependence like on your template, them can monitor it change with 1click
1 Like
@Vadim, @Krey
You both have more than 5 nodes, right?
When you create that %dependence - can you check if smaller nodes have higher % (especially if they are at full capacity) - that would expose one flaw in test data simulation (or deterrent for storage hoarders)
Theory - smaller nodes have higher % Egress per TB stored - meaning it is not worth to grow your node beyond certain level (ignoring the fact that you have to factor in multiple internet connections costs per each smaller node or spin multiple small nodes when one fills up)
I will check this, but most of my node is 4TB, and 2-3 are 2TB, 8tb nodes not full yet.
1 Like
if anyone is or has making a capture/log script for the information we post then by all means share it…
im getting tired of dealing with it… but i think the layout is pretty good tho maybe we want to add daily deletions and then do a full summery of the month and the current months.
maybe add a max possible node size based on deletions and ingress or maybe a monthly growth ratio
1 Like