Portál AbcLinuxu, 6. května 2025 16:41
Mam
HPE ProLiant MicroServer Gen8 12GB RAM (SSD na system a plotnove disky do RAIDu) + Openmediavault + ZFS
a velka spokojenost.
V dnesni dobe bych asi zkusil FreeNAS misto Openmediavault.
D.
# fio --name=seqwrite --rw=write --direct=1 --ioengine=libaio --bs=10M --numjobs=4 --size=3G --runtime=30 --group_reporting seqwrite: (g=0): rw=write, bs=(R) 10.0MiB-10.0MiB, (W) 10.0MiB-10.0MiB, (T) 10.0MiB-10.0MiB, ioengine=libaio, iodepth=1 ... fio-3.12 Starting 4 processes seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) Jobs: 4 (f=4): [W(4)][100.0%][w=340MiB/s][w=34 IOPS][eta 00m:00s] seqwrite: (groupid=0, jobs=4): err= 0: pid=2827: Fri Jul 2 09:20:12 2021 write: IOPS=35, BW=351MiB/s (368MB/s)(10.3GiB/30062msec); 0 zone resets slat (usec): min=1301, max=284189, avg=5095.55, stdev=9403.91 clat (msec): min=10, max=916, avg=108.87, stdev=95.46 lat (msec): min=13, max=1084, avg=113.97, stdev=97.74 clat percentiles (msec): | 1.00th=[ 15], 5.00th=[ 25], 10.00th=[ 30], 20.00th=[ 37], | 30.00th=[ 47], 40.00th=[ 58], 50.00th=[ 75], 60.00th=[ 103], | 70.00th=[ 136], 80.00th=[ 171], 90.00th=[ 230], 95.00th=[ 292], | 99.00th=[ 376], 99.50th=[ 447], 99.90th=[ 885], 99.95th=[ 919], | 99.99th=[ 919] bw ( KiB/s): min=20439, max=204800, per=25.37%, avg=91068.26, stdev=31490.80, samples=236 iops : min= 1, max= 20, avg= 8.75, stdev= 3.11, samples=236 lat (msec) : 20=2.37%, 50=30.74%, 100=26.38%, 250=32.16%, 500=7.97% lat (msec) : 1000=0.38% cpu : usr=1.66%, sys=1.05%, ctx=6872, majf=0, minf=52 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued rwts: total=0,1054,0,0 short=0,0,0,0 dropped=0,0,0,0 latency : target=0, window=0, percentile=100.00%, depth=1 Run status group 0 (all jobs): WRITE: bw=351MiB/s (368MB/s), 351MiB/s-351MiB/s (368MB/s-368MB/s), io=10.3GiB (11.1GB), run=30062-30062msec Disk stats (read/write): drbd0: ios=1/23989, merge=0/0, ticks=76/2306840, in_queue=2312500, util=99.49%, aggrios=1/23991, aggrmerge=0/0, aggrticks=4/859528, aggrin_queue=860172, aggrutil=96.89% zd0: ios=1/23991, merge=0/0, ticks=4/859528, in_queue=860172, util=96.89%bs=64k
# fio --name=seqwrite --rw=write --direct=1 --ioengine=libaio --bs=64k --numjobs=4 --size=3G --runtime=30 --group_reporting seqwrite: (g=0): rw=write, bs=(R) 64.0KiB-64.0KiB, (W) 64.0KiB-64.0KiB, (T) 64.0KiB-64.0KiB, ioengine=libaio, iodepth=1 ... fio-3.12 Starting 4 processes seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) Jobs: 4 (f=4): [W(4)][100.0%][w=139MiB/s][w=2229 IOPS][eta 00m:00s] seqwrite: (groupid=0, jobs=4): err= 0: pid=20925: Fri Jul 2 09:24:44 2021 write: IOPS=2138, BW=134MiB/s (140MB/s)(4011MiB/30002msec); 0 zone resets slat (usec): min=28, max=186285, avg=66.17, stdev=983.59 clat (usec): min=5, max=847636, avg=1799.67, stdev=8268.41 lat (usec): min=356, max=871106, avg=1866.48, stdev=8731.90 clat percentiles (usec): | 1.00th=[ 371], 5.00th=[ 424], 10.00th=[ 469], 20.00th=[ 529], | 30.00th=[ 603], 40.00th=[ 783], 50.00th=[ 1582], 60.00th=[ 1876], | 70.00th=[ 2147], 80.00th=[ 2442], 90.00th=[ 3752], 95.00th=[ 4146], | 99.00th=[ 5800], 99.50th=[ 10814], 99.90th=[ 18220], 99.95th=[ 19530], | 99.99th=[633340] bw ( KiB/s): min= 3968, max=41728, per=25.41%, avg=34780.64, stdev=6360.12, samples=235 iops : min= 62, max= 652, avg=543.37, stdev=99.36, samples=235 lat (usec) : 10=0.01%, 20=0.01%, 50=0.01%, 100=0.01%, 250=0.03% lat (usec) : 500=14.79%, 750=24.03%, 1000=5.51% lat (msec) : 2=21.33%, 4=27.47%, 10=6.26%, 20=0.53%, 50=0.02% lat (msec) : 100=0.01%, 250=0.01%, 750=0.01%, 1000=0.01% cpu : usr=1.01%, sys=4.35%, ctx=65714, majf=0, minf=55 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued rwts: total=0,64173,0,0 short=0,0,0,0 dropped=0,0,0,0 latency : target=0, window=0, percentile=100.00%, depth=1 Run status group 0 (all jobs): WRITE: bw=134MiB/s (140MB/s), 134MiB/s-134MiB/s (140MB/s-140MB/s), io=4011MiB (4206MB), run=30002-30002msec Disk stats (read/write): drbd0: ios=1/69350, merge=0/0, ticks=0/1137592, in_queue=1137612, util=99.29%, aggrios=1/69393, aggrmerge=0/0, aggrticks=0/103032, aggrin_queue=103052, aggrutil=81.78% zd0: ios=1/69393, merge=0/0, ticks=0/103032, in_queue=103052, util=81.78%
volblocksize
) som musel nastavit na 128k, inac to bolo velmi pomale. Kompresiu mam vypnutu, primarycache=metadata
a secondarycache=metadata
- toto mi myslim tiez pomohlo.
Zapnuta kompresia (lz4
):
# fio --name=seqwrite --rw=write --direct=1 --ioengine=libaio --bs=10M --numjobs=4 --size=3G --runtime=30 --group_reporting seqwrite: (g=0): rw=write, bs=(R) 10.0MiB-10.0MiB, (W) 10.0MiB-10.0MiB, (T) 10.0MiB-10.0MiB, ioengine=libaio, iodepth=1 ... fio-3.12 Starting 4 processes seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) seqwrite: Laying out IO file (1 file / 3072MiB) Jobs: 3 (f=3): [W(3),_(1)][90.3%][w=380MiB/s][w=38 IOPS][eta 00m:03s] seqwrite: (groupid=0, jobs=4): err= 0: pid=5138: Fri Jul 2 09:30:52 2021 write: IOPS=42, BW=428MiB/s (449MB/s)(11.0GiB/28701msec); 0 zone resets slat (usec): min=982, max=63277, avg=4766.85, stdev=3506.68 clat (msec): min=8, max=646, avg=85.36, stdev=87.74 lat (msec): min=12, max=648, avg=90.14, stdev=87.62 clat percentiles (msec): | 1.00th=[ 11], 5.00th=[ 14], 10.00th=[ 17], 20.00th=[ 33], | 30.00th=[ 42], 40.00th=[ 48], 50.00th=[ 55], 60.00th=[ 64], | 70.00th=[ 77], 80.00th=[ 106], 90.00th=[ 243], 95.00th=[ 305], | 99.00th=[ 359], 99.50th=[ 380], 99.90th=[ 634], 99.95th=[ 651], | 99.99th=[ 651] bw ( KiB/s): min=20398, max=397319, per=25.48%, avg=111653.06, stdev=52861.71, samples=219 iops : min= 1, max= 38, avg=10.61, stdev= 5.12, samples=219 lat (msec) : 10=0.41%, 20=12.13%, 50=30.86%, 100=35.83%, 250=11.24% lat (msec) : 500=9.36%, 750=0.16% cpu : usr=1.56%, sys=1.12%, ctx=13276, majf=0, minf=47 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued rwts: total=0,1228,0,0 short=0,0,0,0 dropped=0,0,0,0 latency : target=0, window=0, percentile=100.00%, depth=1 Run status group 0 (all jobs): WRITE: bw=428MiB/s (449MB/s), 428MiB/s-428MiB/s (449MB/s-449MB/s), io=11.0GiB (12.9GB), run=28701-28701msec Disk stats (read/write): drbd0: ios=1/26500, merge=0/0, ticks=0/1655876, in_queue=1655352, util=99.55%, aggrios=1/26609, aggrmerge=0/0, aggrticks=0/555608, aggrin_queue=567840, aggrutil=81.35% zd0: ios=1/26609, merge=0/0, ticks=0/555608, in_queue=567840, util=81.35%Ale to sa mi nezda. fio generuje prazdne bloky?
Já jsem OMV zkoušel před cca třemi lety, ale bylo to tako porůznu rozbité (padající a selhávající GUI).
Tiskni
Sdílej:
ISSN 1214-1267, (c) 1999-2007 Stickfish s.r.o.