Prev: [PATCH] slub: Potential stack overflow
Next: [rfc patch] wm8994: range checking issue
From: Justin Piszcz on 24 Mar 2010 08:00
Hi,

2.6.33.1/x86_64 on an e6700 processor:

If you accidentally check [*] for Multi-core processing, the speed is
about 5MiB/s and there are some 254 processes.

When its not enabled, the speed goes back to normal and there are no longer
200-300 processes running, I have reproduced this on multiple machines now.

Enable this option in the kernel, boot the new kernel, do some raid i/o
operations/or a check/resync to the array and it will be 1/4 the speed when
multi-core processing is enabled.
| | [ ] RAID-4/RAID-5/RAID-6 Multicore processing (EXPERIMENTAL| |

With it disabled, we're good:
md0 : active raid6 sdg1[5] sde1[3] sdf1[4] sdd1[2] sdc1[1] sdb1[0]
1562834944 blocks level 6, 64k chunk, algorithm 2 [6/6] [UUUUUU]
[========>............] resync = 44.5% (173910976/390708736) finish=178.3min speed=20253K/sec

Justin.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: [PATCH] slub: Potential stack overflow
Next: [rfc patch] wm8994: range checking issue