Tomasz Chmielewski, on 04/06/2009 10:27 PM wrote: > Vladislav Bolkhovitin schrieb: > >>> Encrypted device was created with the following additional options >>> passed to cryptsetup >>> (it provides the most performance on systems where CPU is a >>> bottleneck, but with decreased >>> security when compared to default options): >>> >>> -c aes-ecb-plain -s 128 >>> >>> >>> Generally, CPU on the target was a bottleneck, so I also tested the >>> load on target. >>> >>> >>> md0, crypt columns - averages from dd >>> us, sy, id, wa - averages from vmstat >>> >>> >>> 1. Disk speeds on the target >>> >>> Raw performance: 102.17 MB/s >>> Raw performance (encrypted): 50.21 MB/s >>> >>> >>> 2. Read-ahead on the initiator: 256 (default); md0, crypt - MB/s >>> >>> md0 us sy id wa | crypt us sy id >>> wa STGT 50.63 4% 45% 18% 33% | 32.52 3% 62% >>> 16% 19% >>> SCST (debug + no patches) 43.75 0% 26% 30% 44% | 42.05 0% 84% 1% >>> 15% >>> SCST (fullperf + patches) 45.18 0% 25% 33% 42% | 44.12 0% 81% 2% >>> 17% >>> >>> >>> 3. Read-ahead on the initiator: 16384; md0, crypt - MB/s >>> >>> md0 us sy id wa | crypt us sy id >>> wa STGT 56.43 3% 55% 2% 40% | 46.90 3% >>> 90% 3% 4% >>> SCST (debug + no patches) 73.85 0% 58% 1% 41% | 42.70 0% 85% 0% >>> 15% >>> SCST (fullperf + patches) 76.27 0% 63% 1% 36% | 42.52 0% 85% 0% >>> 15% >> Good! You proved that: >> >> 1. SCST is capable to work much better than STGT: 35% for md and 37% for >> crypt considering maximum values. >> >> 2. Default read-ahead size isn't appropriate for remote data access >> cases and should be increased. I slowly have been discussing it in past >> few months with Wu Fengguang, the read-ahead maintainer. > > Note that crypt performance for SCST was worse than that of STGT for > large read-ahead values. > Also, SCST performance on crypt device was more or less the same with > 256 and 16384 readahead values. I wonder why performance didn't increase > here while increasing readahead values? This is a very big topic. In short, increasing RA alone isn't sufficient, because, while the bigger value transferred over the uplink, the backend storage can get rotated too far, so, to continue reading data from it, there will be a need to wait for that rotation completed. Try together with the RA increase also decrease max_sectors_kb to 128K or even to 64K. Also, the above actions done on the target can also be quite positive. > Could anyone recheck if it's the same on some other system? > >> Which IO scheduler on the target did you use? I guess, deadline? If so, >> you should try with CFQ as well. > > I used CFQ. You didn't apply io_context-XXX.patch, correct? With it you should see a noticeable increase, like in http://scst.sourceforge.net/vl_res.txt. -- To unsubscribe from this list: send the line "unsubscribe stgt" in the body of a message to majordomo at vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html |