| === synapt is now known as nate | ||
| faekjarz | Applied logic reinforces your points! However, i still intend to benchmark my swap. Perhaps we should look at this issue, through the glasses of science, and come up with creative solutions? | 00:02 |
|---|---|---|
| faekjarz | well, perhaps we can come up with recommendations for alternative channels | 00:07 |
| faekjarz | if the ubuntu-server channel lacks the XP, there's probably a list of other channels, you'd like to recommend (this is not a diss, it's a sci request) | 00:11 |
| sarnold | I suggest returning to the question when you're not drunk | 00:12 |
| faekjarz | sarnold: your advice confidence level has always been quite high/desirable! (thank you, again) | 00:16 |
| sarnold | :D | 00:16 |
| faekjarz | ;) | 00:16 |
| faekjarz | How obnoxious is my being persistent? Here's my maths: the NVMe SSD is Gen3 x4, via B550 chipset (Gen3 x4 link to the CPU). That SerDes phy achieves less than "cryptsetup benchmark" AES 512b XTS throughput - that's ~3.9gig vs. 4.1gig per second. With ext4 overhead. Does it even matter? | 00:36 |
| faekjarz | (i think i'm not drunk enough to be be incompetent.) | 00:39 |
| faekjarz | sure, i can maths it out myself, but i'd like to read your XP | 00:41 |
| sarnold | if you're doing 4gbps of swap traffic you need more ram | 00:58 |
| === chris14_ is now known as chris14 | ||
| === shokohsc51087 is now known as shokohsc5108 | ||
| === shokohsc51086 is now known as shokohsc5108 | ||
| === shokohsc51084 is now known as shokohsc5108 | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!