Post B2fbnVbZcNLXf3gfWS by ryanc@infosec.exchange
 (DIR) More posts by ryanc@infosec.exchange
 (DIR) Post #B2eVvjmxyOwN5Kb3mS by ryanc@infosec.exchange
       2026-01-25T19:02:23Z
       
       1 likes, 0 repeats
       
       # zpool list -v tankNAME                      SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOTtank                     89.0T  15.1T  73.9T        -         -     0%    16%  1.00x    ONLINE  -  raidz3-0               87.3T  14.9T  72.4T        -         -     0%  17.1%      -    ONLINE    crypt-slot-l0        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-r0        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-l1        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-r1        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-l2        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-r2        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-l3        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-r3        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-l4        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-r4        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-l5        7.28T      -      -        -         -      -      -      -    ONLINE    crypt-slot-r5        7.28T      -      -        -         -      -      -      -    ONLINEspecial                      -      -      -        -         -      -      -      -         -  mirror-2               1.64T   121G  1.52T        -         -     0%  7.21%      -    ONLINE    crypt-slot-cd-part4  1.65T      -      -        -         -      -      -      -    ONLINE    crypt-slot-cu-part4  1.65T      -      -        -         -      -      -      -    ONLINE    crypt-slot-cl-part4  1.65T      -      -        -         -      -      -      -    ONLINElogs                         -      -      -        -         -      -      -      -         -  mirror-1               31.5G   188K  31.5G        -         -     0%  0.00%      -    ONLINE    crypt-slot-cd-part3  32.0G      -      -        -         -      -      -      -    ONLINE    crypt-slot-cu-part3  32.0G      -      -        -         -      -      -      -    ONLINE    crypt-slot-cl-part3  32.0G      -      -        -         -      -      -      -    ONLINE
       
 (DIR) Post #B2eW0vuCLq3I70SNl2 by louis@social.louis-vallat.dev
       2026-01-25T19:04:25.459Z
       
       0 likes, 0 repeats
       
       @ryanc@infosec.exchange crypt-slot? ​:neocat_think:​
       
 (DIR) Post #B2eXmh58rrDZzYui5Q by ryanc@infosec.exchange
       2026-01-25T19:17:29Z
       
       1 likes, 0 repeats
       
       @louis the disks are all using LUKS
       
 (DIR) Post #B2eY1r61uuSZOWYgyW by louis@social.louis-vallat.dev
       2026-01-25T19:26:59.610Z
       
       0 likes, 0 repeats
       
       @ryanc@infosec.exchange so not giving ZFS raw access to the disks? ​:distraught:​ what about ZFS encryption? ​:neocat_pleading:​
       
 (DIR) Post #B2eu1P04ZjspFPxeme by ryanc@infosec.exchange
       2026-01-25T19:53:04Z
       
       1 likes, 0 repeats
       
       @louis ZFS encryption performance is bad.
       
 (DIR) Post #B2eu1QOvMpM9amb19s by louis@social.louis-vallat.dev
       2026-01-25T23:33:24.028Z
       
       0 likes, 0 repeats
       
       @ryanc@infosec.exchange even recently (and on modern hardware)? I haven't seen any performance (latency or throughout) degradation on SSDs or HDDs on modern-ish hardware ​:noted:​
       
 (DIR) Post #B2eyB0uJPHYfEm6aH2 by ryanc@infosec.exchange
       2026-01-26T00:19:45Z
       
       1 likes, 0 repeats
       
       @louis It's on an older Atom CPU that doesn't have AVX instruction set support, so it's a bit shit. It's fine on my laptop with a recent CPU.
       
 (DIR) Post #B2eyEDqHlra0Dukg7s by louis@social.louis-vallat.dev
       2026-01-26T00:20:33.807Z
       
       0 likes, 0 repeats
       
       @ryanc@infosec.exchange aaaaaaaaaah gotcha'! ​:neocat_cool_fingerguns:​ makes more sense then (although a bit cursed) ​:bleh:​
       
 (DIR) Post #B2fbnVbZcNLXf3gfWS by ryanc@infosec.exchange
       2026-01-26T00:55:00Z
       
       1 likes, 0 repeats
       
       @louis It has a QAT coprocessor that supports AES-GCM, but that's still a bottleneck for some reason.
       
 (DIR) Post #B2vVbBh0Kqk7EWKOLg by GermanoMassullo@mastodon.social
       2026-02-02T20:02:51Z
       
       0 likes, 0 repeats
       
       @ryanc @louis nope
       
 (DIR) Post #B2vVbCi6YFDWODqmhc by ryanc@infosec.exchange
       2026-02-02T21:31:56Z
       
       1 likes, 0 repeats
       
       @louis @GermanoMassullo I have receipts.
       
 (DIR) Post #B2vVbI8AOm2FCV5AWG by ryanc@infosec.exchange
       2026-02-02T21:33:37Z
       
       1 likes, 0 repeats
       
       @louis @GermanoMassullo less flippantly, the performance is shit on the box I need it to run on