[ilyt]

I guess for just a dashboard it doesn't matter either way, it's just running a screen; I added it coz I was worried that say background package upgrade would be stopped midway and in worst scenario, midway to upgrading kernel.

>If you're talking about the degradation that accompanies any heavy use of flash NAND then i'm not sure what to tell you other than it can be mostly mitigated in software, and that it's a universal trait; RPis are by no means the only system to suffer from those issues.

I'm not but any brand of SD card to recommend ? I was using some samsung EVO (bought off a store, not random e-bay seller) and my cluster pretty much self-destroyed even tho I severly limited writes in the first place.

They also failed in weird way where writes were confirmed but never landed on flash so the system worked till it had to remove latest writes from buffer memory... then it said FS is corrupted and after reboot it was back to previous state...1