1 |
Shouldn't you be using sdparm, and not hdparm for your sata drives? |
2 |
|
3 |
On 1/23/06, gentoo-performance+help@g.o < |
4 |
gentoo-performance+help@g.o> wrote: |
5 |
> |
6 |
> Topics (messages 32 throught 61): |
7 |
> |
8 |
> [gentoo-performance] |
9 |
> 32 - ??? <bolotov-bag@×××××××.ru> |
10 |
> |
11 |
> [gentoo-performance] gentoo-performance |
12 |
> 33 - Ken Robbins <gatliffe@×××××.com> |
13 |
> |
14 |
> [gentoo-performance] gentoo-performance |
15 |
> 34 - lnxg33k <lnxg33k@×××××.com> |
16 |
> |
17 |
> [gentoo-performance] gentoo-performance |
18 |
> 35 - Chris <chris@×××××××××××.net> |
19 |
> |
20 |
> [gentoo-performance] gentoo-performance |
21 |
> 36 - lnxg33k <lnxg33k@×××××.com> |
22 |
> |
23 |
> [gentoo-performance] gentoo-performance |
24 |
> 37 - Alex Efros <powerman@××××××××××××××××××.com> |
25 |
> |
26 |
> [gentoo-performance] gentoo-performance |
27 |
> 38 - Kyle Lutze <kyle@×××××××××××.com> |
28 |
> |
29 |
> [gentoo-performance] gentoo-performance |
30 |
> 39 - Christopher Bergström <cbergstrom@×××××××××.com> |
31 |
> |
32 |
> [gentoo-performance] gentoo-performance |
33 |
> 40 - Jeremy Brake <gentoolists@×××××××××××.nz> |
34 |
> |
35 |
> [gentoo-performance] gentoo-performance |
36 |
> 41 - lnxg33k <lnxg33k@×××××.com> |
37 |
> |
38 |
> [gentoo-performance] gentoo-performance |
39 |
> 42 - darren kirby <bulliver@×××××××××××.org> |
40 |
> |
41 |
> [gentoo-performance] gentoo-performance |
42 |
> 43 - Michael Liesenfelt <mliesenf@×××××××××.edu> |
43 |
> |
44 |
> [gentoo-performance] gentoo-performance |
45 |
> 44 - Christopher Bergström <cbergstrom@×××××××××.com> |
46 |
> |
47 |
> [gentoo-performance] gentoo-performance |
48 |
> 45 - Alec Warner <warnera6@×××××××.edu> |
49 |
> |
50 |
> [gentoo-performance] gentoo-performance |
51 |
> 46 - Jeremy Brake <gentoolists@×××××××××××.nz> |
52 |
> |
53 |
> [gentoo-performance] gentoo-performance |
54 |
> 47 - darren kirby <bulliver@×××××××××××.org> |
55 |
> |
56 |
> [gentoo-performance] unsubscribe |
57 |
> 48 - "Matthew Coulson" <MCoulson@××××××××××××××××××××××××.uk> |
58 |
> 53 - Alden Huang <alden.huang@×××××.com> |
59 |
> 55 - Geisel Sierote <geisel@×××××××.br> |
60 |
> |
61 |
> [gentoo-performance] gentoo-performance |
62 |
> 49 - Bill Roberts <billbalt@×××××××××××××.com> |
63 |
> |
64 |
> [gentoo-performance] gentoo-performance |
65 |
> 50 - Christopher Bergström <cbergstrom@×××××××××.com> |
66 |
> |
67 |
> [gentoo-performance] gentoo-performance (sync speedups) |
68 |
> 52 - lnxg33k <lnxg33k@×××××.com> |
69 |
> |
70 |
> [gentoo-performance] gentoo-performance@l.g.o |
71 |
> 54 - checl <check@×××××××××.cx> |
72 |
> |
73 |
> [gentoo-performance] How to unsubscribe |
74 |
> 56 - Oopsz <oopsz@××××××××××.com> |
75 |
> |
76 |
> [gentoo-performance] gentoo-performance (sync speedups) |
77 |
> 57 - Alec Warner <warnera6@×××××××.edu> |
78 |
> |
79 |
> [gentoo-performance] gentoo-performance (sync speedups) |
80 |
> 58 - lnxg33k <lnxg33k@×××××.com> |
81 |
> |
82 |
> [gentoo-performance] How to get Maximum performance in Graphics on Nvidia |
83 |
> Drivers |
84 |
> 59 - XFry <xfry@××××××.ru> |
85 |
> |
86 |
> |
87 |
> |
88 |
> |
89 |
> |
90 |
> |
91 |
> Hi my first gentoo performance came today but it way only a header no body |
92 |
> what up with that? |
93 |
> |
94 |
> |
95 |
> *Powered by Gentoo Linux* |
96 |
> Anything free is worth saving up for-Shadow the cat |
97 |
> |
98 |
> ------------------------------ |
99 |
> Yahoo! Photos |
100 |
> Ring in the New Year with Photo Calendars<http://us.rd.yahoo.com/mail_us/taglines/photos/*http://pa.yahoo.com/*http://us.rd.yahoo.com/mail_us/taglines/photos/evt=38087/*http://pg.photos.yahoo.com/ph//page?.file=calendar_splash.html&.dir=>. |
101 |
> Add photos, events, holidays, whatever. |
102 |
> |
103 |
> |
104 |
> Ken Robbins wrote: |
105 |
> > Hi my first gentoo performance came today but it way only a header no |
106 |
> body what up with that? |
107 |
> > |
108 |
> <snip> |
109 |
> |
110 |
> Mine too and I've been listening for a while. I think this ML may be dead |
111 |
> ... |
112 |
> *pokes the darkness* |
113 |
> -- |
114 |
> gentoo-performance@g.o mailing list |
115 |
> |
116 |
> funny that a "high performance linux" has a dead performance ML... LOL |
117 |
> |
118 |
> |
119 |
> lnxg33k wrote: |
120 |
> |
121 |
> > Ken Robbins wrote: |
122 |
> > |
123 |
> >> Hi my first gentoo performance came today but it way only a header no |
124 |
> >> body what up with that? |
125 |
> > |
126 |
> > <snip> |
127 |
> > |
128 |
> > Mine too and I've been listening for a while. I think this ML may be |
129 |
> > dead ... *pokes the darkness* |
130 |
> |
131 |
> -- |
132 |
> gentoo-performance@g.o mailing list |
133 |
> |
134 |
> Chris wrote: |
135 |
> > funny that a "high performance linux" has a dead performance ML... LOL |
136 |
> <snip> |
137 |
> |
138 |
> Could be evidence that the "ricer" crowd doesn't read? (i.e. they post to |
139 |
> more |
140 |
> generic lists or use other mediums instead of something specific for their |
141 |
> needs) |
142 |
> -- |
143 |
> gentoo-performance@g.o mailing list |
144 |
> |
145 |
> Hi! |
146 |
> |
147 |
> On Fri, Jan 20, 2006 at 03:30:29AM +0100, Chris wrote: |
148 |
> > > Mine too and I've been listening for a while. I think this ML may be |
149 |
> > > dead ... *pokes the darkness* |
150 |
> > funny that a "high performance linux" has a dead performance ML... LOL |
151 |
> |
152 |
> It isn't dead because a lot of attentive listeners subscribed to it... :-) |
153 |
> |
154 |
> -- |
155 |
> WBR, Alex. |
156 |
> -- |
157 |
> gentoo-performance@g.o mailing list |
158 |
> |
159 |
> lnxg33k wrote: |
160 |
> > Chris wrote: |
161 |
> > |
162 |
> >> funny that a "high performance linux" has a dead performance ML... LOL |
163 |
> > |
164 |
> > <snip> |
165 |
> > |
166 |
> > Could be evidence that the "ricer" crowd doesn't read? (i.e. they post |
167 |
> > to more generic lists or use other mediums instead of something specific |
168 |
> > for their needs) |
169 |
> |
170 |
> Time to throw in a performance info post. |
171 |
> |
172 |
> Has anybody done any tests on the different between the one core and |
173 |
> dual core opteron processors? I currently have an opteron 242 and a gig |
174 |
> of ram on a tyan mobo, and I'm curious as to which would allow me to |
175 |
> compile programs faster, as I lend the systems out to a lot of groups |
176 |
> that have slow systems |
177 |
> |
178 |
> Kyle |
179 |
> -- |
180 |
> gentoo-performance@g.o mailing list |
181 |
> |
182 |
> lnxg33k wrote: |
183 |
> |
184 |
> > Chris wrote: |
185 |
> > |
186 |
> >> funny that a "high performance linux" has a dead performance ML... LOL |
187 |
> > |
188 |
> > <snip> |
189 |
> > |
190 |
> > Could be evidence that the "ricer" crowd doesn't read? (i.e. they post |
191 |
> > to more generic lists or use other mediums instead of something |
192 |
> > specific for their needs) |
193 |
> |
194 |
> Since we're all here and saying hello.. Someone have a performance |
195 |
> question or a good tip to add to the list? |
196 |
> |
197 |
> The one thing that first pops to my head is some hdparm results I've had |
198 |
> and if maybe it's either my kernel setup or how I'm testing with hdparm... |
199 |
> |
200 |
> Anyhow.. |
201 |
> |
202 |
> Raptors on SATA (Model Number: WDC WD360GD-00FNA0) (-c3 -u1 -A1 |
203 |
> -W1 -d1 -a256 -M254 -m16 -X70 ) |
204 |
> Kernel config |
205 |
> CONFIG_SCSI_SATA_SIL=y |
206 |
> |
207 |
> There's an alternative driver, but haven't tested it... |
208 |
> |
209 |
> hdparm -tT /dev/hde |
210 |
> |
211 |
> /dev/hde: |
212 |
> Timing cached reads: 956 MB in 2.00 seconds = 477.02 MB/sec |
213 |
> Timing buffered disk reads: 116 MB in 3.03 seconds = 38.26 MB/sec |
214 |
> |
215 |
> hdparm -tT --direct /dev/hde |
216 |
> |
217 |
> /dev/hde: |
218 |
> Timing O_DIRECT cached reads: 244 MB in 2.01 seconds = 121.26 MB/sec |
219 |
> Timing O_DIRECT disk reads: 124 MB in 3.01 seconds = 41.17 MB/sec |
220 |
> |
221 |
> Laptop 7k60 (Model Number: HTS726060M9AT00) |
222 |
> hdparm -tT /dev/hdc |
223 |
> |
224 |
> /dev/hdc: |
225 |
> Timing cached reads: 1980 MB in 2.00 seconds = 989.94 MB/sec |
226 |
> Timing buffered disk reads: 118 MB in 3.04 seconds = 38.81 MB/sec |
227 |
> |
228 |
> |
229 |
> hdparm -tT --direct /dev/hdc |
230 |
> |
231 |
> /dev/hdc: |
232 |
> Timing O_DIRECT cached reads: 364 MB in 2.02 seconds = 180.54 MB/sec |
233 |
> Timing O_DIRECT disk reads: 120 MB in 3.04 seconds = 39.42 MB/sec |
234 |
> |
235 |
> |
236 |
> I also can't set the raptors into UDMA6 tried -X70 with no luck.. |
237 |
> |
238 |
> Any suggestions? |
239 |
> |
240 |
> Thanks |
241 |
> |
242 |
> C. |
243 |
> -- |
244 |
> gentoo-performance@g.o mailing list |
245 |
> |
246 |
> How about speeding up the wait time on updating the portage cache after |
247 |
> a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
248 |
> through.. |
249 |
> are there any known ways to "vrrmmm" this up a little? |
250 |
> |
251 |
> Jeremy |
252 |
> -- |
253 |
> gentoo-performance@g.o mailing list |
254 |
> |
255 |
> Jeremy Brake wrote: |
256 |
> > How about speeding up the wait time on updating the portage cache after |
257 |
> > a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
258 |
> > through.. |
259 |
> > are there any known ways to "vrrmmm" this up a little? |
260 |
> > |
261 |
> > Jeremy |
262 |
> |
263 |
> Although not exactly what you're asking, you might want to look at |
264 |
> RSYNC_EXCLUDEFORM. Cuts down on what is checked during rsync. I assume |
265 |
> that |
266 |
> this could also cut down on the cache update time since there would be |
267 |
> less to |
268 |
> check? Also cuts down on the amount of space eaten up by portage. |
269 |
> -- |
270 |
> gentoo-performance@g.o mailing list |
271 |
> |
272 |
> quoth the Jeremy Brake: |
273 |
> > How about speeding up the wait time on updating the portage cache after |
274 |
> > a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
275 |
> > through.. |
276 |
> > are there any known ways to "vrrmmm" this up a little? |
277 |
> |
278 |
> +1 |
279 |
> |
280 |
> Mine sped up for all of a day, but is slow as molasses once again. If I |
281 |
> did my |
282 |
> syncs during the day it might even peeve me... |
283 |
> |
284 |
> > Jeremy |
285 |
> |
286 |
> -d |
287 |
> -- |
288 |
> darren kirby :: Part of the problem since 1976 :: http://badcomputer.org |
289 |
> "...the number of UNIX installations has grown to 10, with more |
290 |
> expected..." |
291 |
> - Dennis Ritchie and Ken Thompson, June 1972 |
292 |
> |
293 |
> |
294 |
> lnxg33k wrote: |
295 |
> |
296 |
> > Although not exactly what you're asking, you might want to look at |
297 |
> > RSYNC_EXCLUDEFORM. Cuts down on what is checked during rsync. I assume |
298 |
> > that this could also cut down on the cache update time since there |
299 |
> > would be less to check? Also cuts down on the amount of space eaten up |
300 |
> > by portage. |
301 |
> |
302 |
> |
303 |
> I'll second RSYNC exclusions. It really does speed up syncing especially |
304 |
> on headless servers which don't need x11-*/*. |
305 |
> |
306 |
> -- |
307 |
> Michael Liesenfelt |
308 |
> University of Florida |
309 |
> Innovative Nuclear Space Power and Propulsion Institute |
310 |
> |
311 |
> |
312 |
> |
313 |
> darren kirby wrote: |
314 |
> |
315 |
> quoth the Jeremy Brake: |
316 |
> |
317 |
> How about speeding up the wait time on updating the portage cache after |
318 |
> a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
319 |
> through.. |
320 |
> are there any known ways to "vrrmmm" this up a little? |
321 |
> |
322 |
> +1 |
323 |
> |
324 |
> Mine sped up for all of a day, but is slow as molasses once again. If I did my |
325 |
> syncs during the day it might even peeve me... |
326 |
> |
327 |
> What kind of hardware are you guys running on? My laptop isn't on cron |
328 |
> and I do it every couple days or so and it finishes in around 15-30 |
329 |
> minutes.. I've never really paid any attention.. How long is yours taking? |
330 |
> |
331 |
> What I am curious about is... what's it really doing when it says 51-52%.. |
332 |
> That 1% seems to take forever.. |
333 |
> |
334 |
> C. |
335 |
> -- gentoo-performance@g.o mailing list |
336 |
> -----BEGIN PGP SIGNED MESSAGE----- |
337 |
> Hash: SHA1 |
338 |
> |
339 |
> Jeremy Brake wrote: |
340 |
> > How about speeding up the wait time on updating the portage cache after |
341 |
> > a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
342 |
> > through.. |
343 |
> > are there any known ways to "vrrmmm" this up a little? |
344 |
> > |
345 |
> > Jeremy |
346 |
> |
347 |
> Well the current problem is that in the 2.0 portage branch the cache |
348 |
> code sucks. This is fixed in ~arch portage ( the 2.1_pre series ). For |
349 |
> you users that don't want to upgrade to unstable, you should be able to |
350 |
> use cdb to speed up the process. |
351 |
> |
352 |
> Setting RSYNC_EXCLUDES will not speed up the second half of the --sync ( |
353 |
> the --metadata portion ). |
354 |
> |
355 |
> Explanation: The Portage Tree has a ton of ebuilds in it, when you do |
356 |
> something like emerge -pv <blar> portage used to have to go read each |
357 |
> ebuild and be like "oh what are the USE flags, the dependencies, etc.." |
358 |
> This takes a lot of time, especially for something like emerge -uD world |
359 |
> that may touch thousands of packages. |
360 |
> |
361 |
> So to mitigate this issue portage has a "metadata cache" where it stores |
362 |
> the ebuild metadata so it doesn't have to source each ebuild every time. |
363 |
> This gives performance increases ( reportedly ) of 100x speed-ups. |
364 |
> |
365 |
> However, generating the metadata is time consuming, even on decent |
366 |
> hardware it will probably take 30-45 minutes. To not piss the users |
367 |
> off, Gentoo calculates this metadata serve-side and serves it to you |
368 |
> during sync. |
369 |
> |
370 |
> The Rsync portion of emerge --sync is pretty much everything before the |
371 |
> "updating metadata cache". This should be pretty standard as far as |
372 |
> time goes on all boxes. However the second portion is where portage has |
373 |
> to take the generated server-side cache and kind of "meld" it into your |
374 |
> already existant local cache ( stored in /var/cache/edb/dep for those of |
375 |
> you whom are curious ). This didn't used to take a bunch of time..until |
376 |
> KDE split their packages from KDE-monolith to KDE-meta. The X.org |
377 |
> switch probably won't help much in this area either. Particularly the |
378 |
> slowdown at 50-52% is due to this portion of the tree. |
379 |
> |
380 |
> There is a new cache system in the 2.1 branch of portage, or using cdb, |
381 |
> or perhaps even an sqlite back-end (patches and modules are out there), |
382 |
> will help until such time as the 2.1 branch is stablized ( probably not |
383 |
> too soon ). |
384 |
> |
385 |
> - -Alec Warner (antarus) |
386 |
> |
387 |
> Disclaimer: I'm not a developer (yet) but I've worked with the portage |
388 |
> team for some time. Release dates, features, and other things are not |
389 |
> gauranteed to be correct just because I said so :) |
390 |
> -----BEGIN PGP SIGNATURE----- |
391 |
> Version: GnuPG v1.4.2 (GNU/Linux) |
392 |
> Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org |
393 |
> |
394 |
> iQIVAwUBQ9CExWzglR5RwbyYAQLJEhAAj7i2cRVgv1iBfi61mGkn9q1t/K5JEJcv |
395 |
> zK07bTmMDirviessdKiZVSufXdWV79tW0MDEVqmGI9t+V+uwM77aFbge1VSkEaQB |
396 |
> RWetuQL8tQRUX1KvQ+ItScVtbqKIAQe/Jn+BwSim05jF3+fF15Z6EpPPKypNLQxK |
397 |
> 2ef/bcJ91Gctv0xcd6j943uOPFDCt05Ahe06/pQ0wgGdnAcKLOy/RVwDOVfprXim |
398 |
> AwiVsU4aCxRI056RkEj1VuSwSYQEa91WKpTGv81lZVkRW8LRxkgc7UAydMYGjOY5 |
399 |
> 1prjv2Koyly5ycVvxshKVmLfuaqByY9bBnlklyKDdFW1ZD1udCflCLxmz3GuTCsm |
400 |
> /FHce+Y9smqq3sF1wV7DYXu9vTnLdBqVBlDq4cEtd5XdgQm3TJ5rUGF2Mepicyhg |
401 |
> Bg4ibDExDB5eKWCnGU3ioSCd8TY9cdR9ZXxpm6JLGfr9ztog0/vScsIZbj3dS4RH |
402 |
> WqGklvW0F9N6NjP26WJwtQsmSIZpSJU4yPxneZOdQxGOUfdNPQap+qO+rZaitPKW |
403 |
> yWaikaiSuecPSul0ithpGnCFttPHrBLyNKNl7aom04Bht4KSdaqzriIL/RylWzHW |
404 |
> OgazUV9RuVdI8oEx+q3zKzOgvG3dXeL7HNdhH+j9noIyGVa8ouNHWMGfFUc1baxV |
405 |
> 7eyB1buSeQY= |
406 |
> =vWqe |
407 |
> -----END PGP SIGNATURE----- |
408 |
> -- |
409 |
> gentoo-performance@g.o mailing list |
410 |
> |
411 |
> Alec Warner wrote: |
412 |
> |
413 |
> >-----BEGIN PGP SIGNED MESSAGE----- |
414 |
> >Hash: SHA1 |
415 |
> > |
416 |
> >Jeremy Brake wrote: |
417 |
> > |
418 |
> > |
419 |
> >>How about speeding up the wait time on updating the portage cache after |
420 |
> >>a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
421 |
> >>through.. |
422 |
> >>are there any known ways to "vrrmmm" this up a little? |
423 |
> >> |
424 |
> >>Jeremy |
425 |
> >> |
426 |
> >> |
427 |
> > |
428 |
> >Well the current problem is that in the 2.0 portage branch the cache |
429 |
> >code sucks. This is fixed in ~arch portage ( the 2.1_pre series ). For |
430 |
> >you users that don't want to upgrade to unstable, you should be able to |
431 |
> >use cdb to speed up the process. |
432 |
> > |
433 |
> >Setting RSYNC_EXCLUDES will not speed up the second half of the --sync ( |
434 |
> >the --metadata portion ). |
435 |
> > |
436 |
> >Explanation: *snip* |
437 |
> > |
438 |
> > |
439 |
> Thanks Alec, thats a really awesome explaination :) |
440 |
> |
441 |
> My server runs a 5am script which does this, so i'm not too worried |
442 |
> about that machine. For those who are curious, its an Athlon 1800+ on a |
443 |
> 10Mbit link, and it takes between 1 and 10 mins to process " emerge |
444 |
> --sync --quiet; emerge -upvD world; glsa-check -t all " |
445 |
> |
446 |
> My home pc is on a 2Mbit link, so I only sync when i feel like checking |
447 |
> for updates, or when I want to install something new. This will take |
448 |
> minimum of 10 mins just to update the cache most times, sometimes more. |
449 |
> Being a home pc, I'm happy to have some unstable stuff installed. How |
450 |
> messy would it be to just run ~amd64 portage? would this work, or do I |
451 |
> ideally need to make the entire base system ~amd64? (ugh). |
452 |
> |
453 |
> Jeremy |
454 |
> |
455 |
> |
456 |
> -- |
457 |
> gentoo-performance@g.o mailing list |
458 |
> |
459 |
> quoth the Christopher Bergström: |
460 |
> > darren kirby wrote: |
461 |
> > quoth the Jeremy Brake: |
462 |
> > |
463 |
> > How about speeding up the wait time on updating the portage cache after |
464 |
> > a sync.. even on my AMD 64 3500 it takes a number of minutes to chug |
465 |
> > through.. |
466 |
> > are there any known ways to "vrrmmm" this up a little? |
467 |
> > |
468 |
> > |
469 |
> > +1 |
470 |
> > |
471 |
> > Mine sped up for all of a day, but is slow as molasses once again. If I |
472 |
> did |
473 |
> > my syncs during the day it might even peeve me... |
474 |
> > |
475 |
> > What kind of hardware are you guys running on? My laptop isn't on cron |
476 |
> > and I do it every couple days or so and it finishes in around 15-30 |
477 |
> > minutes.. I've never really paid any attention.. How long is yours |
478 |
> taking? |
479 |
> |
480 |
> It doesn't make a difference what hardware. I have 4 boxes that run gentoo |
481 |
> (Athlon 2200, Athlon 3200, Apple G4, Sparc U60) and they all run the |
482 |
> actual |
483 |
> sync quite fast, but the 50%-51% takes from 5 minutes on the 3200 to 30 |
484 |
> minutes on the G4 and U60. |
485 |
> |
486 |
> As I mentioned though, I don't let this bother me as I generally sync |
487 |
> ~4:00am |
488 |
> whilst sleeping. |
489 |
> |
490 |
> > What I am curious about is... what's it really doing when it says |
491 |
> 51-52%.. |
492 |
> > That 1% seems to take forever.. |
493 |
> > |
494 |
> > C. |
495 |
> > -- gentoo-performance@g.o mailing list |
496 |
> -d |
497 |
> -- |
498 |
> darren kirby :: Part of the problem since 1976 :: http://badcomputer.org |
499 |
> "...the number of UNIX installations has grown to 10, with more |
500 |
> expected..." |
501 |
> - Dennis Ritchie and Ken Thompson, June 1972 |
502 |
> |
503 |
> |
504 |
> unsubscribe |
505 |
> |
506 |
> _____________________________________________________________________ |
507 |
> Please Note: This e-mail is confidential and may be protected by law. |
508 |
> This e-mail is intended solely for the named recipient(s). If you receive |
509 |
> this e-mail in error, please destroy the copy in your possession |
510 |
> immediately. Please do not disclose the contents to any other person, use |
511 |
> information contained in it for any purpose, store or copy it. Although |
512 |
> this e-mail and any attachments are believed to be free of any virus, which |
513 |
> might affect a computer system into which it is received and opened, Express |
514 |
> Reinforcements Ltd. can not guarantee this and does not accept |
515 |
> responsibility for any damage resulting from the use of this e-mail. |
516 |
> Copyright in this e-mail and any attachments remains with us. Express |
517 |
> Reinforcements Ltd. Registered in England No.1808624. Head Office: |
518 |
> Fordwater Trading Estate, Ford Road, Chertsey, Surrey KT16 8HG |
519 |
> To Visit our website go to http://www.ExpressReinforcements.co.uk |
520 |
> |
521 |
> This message has been checked for all known viruses by UUNET delivered |
522 |
> through the MessageLabs Virus Control Centre. For further information |
523 |
> visit |
524 |
> http://www.uk.uu.net/products/security/virus/ |
525 |
> |
526 |
> -- |
527 |
> gentoo-performance@g.o mailing list |
528 |
> |
529 |
> > |
530 |
> > The one thing that first pops to my head is some hdparm results I've had |
531 |
> > and if maybe it's either my kernel setup or how I'm testing with |
532 |
> hdparm... |
533 |
> > |
534 |
> > Anyhow.. |
535 |
> > |
536 |
> > Raptors on SATA (Model Number: WDC WD360GD-00FNA0) (-c3 -u1 -A1 |
537 |
> > -W1 -d1 -a256 -M254 -m16 -X70 ) |
538 |
> > Kernel config |
539 |
> > CONFIG_SCSI_SATA_SIL=y |
540 |
> > |
541 |
> > There's an alternative driver, but haven't tested it... |
542 |
> > |
543 |
> > hdparm -tT /dev/hde |
544 |
> > |
545 |
> > /dev/hde: |
546 |
> > Timing cached reads: 956 MB in 2.00 seconds = 477.02 MB/sec |
547 |
> > Timing buffered disk reads: 116 MB in 3.03 seconds = 38.26 MB/sec |
548 |
> > |
549 |
> Your times for the raptor seem awfully slow. |
550 |
> |
551 |
> Here the SATA settings for my kernel and my hdparm times. |
552 |
> |
553 |
> antec ~ # grep SATA /usr/src/linux/.config |
554 |
> # CONFIG_BLK_DEV_IDE_SATA is not set |
555 |
> CONFIG_SCSI_SATA=y |
556 |
> CONFIG_SCSI_SATA_AHCI=y |
557 |
> # CONFIG_SCSI_SATA_SVW is not set |
558 |
> # CONFIG_SCSI_SATA_MV is not set |
559 |
> # CONFIG_SCSI_SATA_NV is not set |
560 |
> # CONFIG_SCSI_SATA_QSTOR is not set |
561 |
> # CONFIG_SCSI_SATA_PROMISE is not set |
562 |
> # CONFIG_SCSI_SATA_SX4 is not set |
563 |
> # CONFIG_SCSI_SATA_SIL is not set |
564 |
> # CONFIG_SCSI_SATA_SIL24 is not set |
565 |
> # CONFIG_SCSI_SATA_SIS is not set |
566 |
> # CONFIG_SCSI_SATA_ULI is not set |
567 |
> # CONFIG_SCSI_SATA_VIA is not set |
568 |
> # CONFIG_SCSI_SATA_VITESSE is not set |
569 |
> CONFIG_SCSI_SATA_INTEL_COMBINED=y |
570 |
> antec ~ # hdparm -tT /dev/md0 |
571 |
> |
572 |
> /dev/md0: |
573 |
> Timing cached reads: 2776 MB in 2.00 seconds = 1387.91 MB/sec |
574 |
> Timing buffered disk reads: 398 MB in 3.00 seconds = 132.48 MB/sec |
575 |
> antec ~ # |
576 |
> |
577 |
> Note that this is a RAID0 with two disks, so the buffered disk read time |
578 |
> is double what you should expect on a normal install. |
579 |
> |
580 |
> Good luck. |
581 |
> |
582 |
> Bill Roberts |
583 |
> |
584 |
> |
585 |
> Bill Roberts wrote: |
586 |
> |
587 |
> The one thing that first pops to my head is some hdparm results I've had |
588 |
> and if maybe it's either my kernel setup or how I'm testing with hdparm... |
589 |
> |
590 |
> Anyhow.. |
591 |
> |
592 |
> Raptors on SATA (Model Number: WDC WD360GD-00FNA0) (-c3 -u1 -A1 |
593 |
> -W1 -d1 -a256 -M254 -m16 -X70 ) |
594 |
> Kernel config |
595 |
> CONFIG_SCSI_SATA_SIL=y |
596 |
> |
597 |
> There's an alternative driver, but haven't tested it... |
598 |
> |
599 |
> hdparm -tT /dev/hde |
600 |
> |
601 |
> /dev/hde: |
602 |
> Timing cached reads: 956 MB in 2.00 seconds = 477.02 MB/sec |
603 |
> Timing buffered disk reads: 116 MB in 3.03 seconds = 38.26 MB/sec |
604 |
> |
605 |
> Your times for the raptor seem awfully slow. |
606 |
> |
607 |
> Here the SATA settings for my kernel and my hdparm times. |
608 |
> |
609 |
> antec ~ # grep SATA /usr/src/linux/.config |
610 |
> # CONFIG_BLK_DEV_IDE_SATA is not set |
611 |
> CONFIG_SCSI_SATA=y |
612 |
> CONFIG_SCSI_SATA_AHCI=y |
613 |
> # CONFIG_SCSI_SATA_SVW is not set |
614 |
> # CONFIG_SCSI_SATA_MV is not set |
615 |
> # CONFIG_SCSI_SATA_NV is not set |
616 |
> # CONFIG_SCSI_SATA_QSTOR is not set |
617 |
> # CONFIG_SCSI_SATA_PROMISE is not set |
618 |
> # CONFIG_SCSI_SATA_SX4 is not set |
619 |
> # CONFIG_SCSI_SATA_SIL is not set |
620 |
> # CONFIG_SCSI_SATA_SIL24 is not set |
621 |
> # CONFIG_SCSI_SATA_SIS is not set |
622 |
> # CONFIG_SCSI_SATA_ULI is not set |
623 |
> # CONFIG_SCSI_SATA_VIA is not set |
624 |
> # CONFIG_SCSI_SATA_VITESSE is not set |
625 |
> CONFIG_SCSI_SATA_INTEL_COMBINED=y |
626 |
> antec ~ # hdparm -tT /dev/md0 |
627 |
> |
628 |
> /dev/md0: |
629 |
> Timing cached reads: 2776 MB in 2.00 seconds = 1387.91 MB/sec |
630 |
> Timing buffered disk reads: 398 MB in 3.00 seconds = 132.48 MB/sec |
631 |
> antec ~ # |
632 |
> |
633 |
> Note that this is a RAID0 with two disks, so the buffered disk read time |
634 |
> is double what you should expect on a normal install. |
635 |
> |
636 |
> I guess I should add that 2nd disk in there then.. Anyhow, I'm bound to |
637 |
> what appears to be a not-so-friendly controller.. |
638 |
> |
639 |
> CONFIG_SCSI_SATA_SIL is not set |
640 |
> |
641 |
> I'm going to change the sata_sil.c file to enable UDMA6 and I've seen some |
642 |
> other minor patches as well.. If it all proves stable.. Maybe someone will |
643 |
> actually get it sent upstream. What's further.. I'm on hardened.. So I think |
644 |
> you might have some kernel config options I don't have.. I have to look |
645 |
> again.. |
646 |
> |
647 |
> Cheers, |
648 |
> |
649 |
> C. |
650 |
> -- gentoo-performance@g.o mailing list |
651 |
> Thanks Alec Warner for the great explanation. It still seems like by not |
652 |
> having |
653 |
> portions of the tree by using EXCLUDEFORM and deleting the local dirs that |
654 |
> you'd save some time in the --metadata part of sync as less ebuilds are |
655 |
> available to be checked. Is this simply a wrong misconception? |
656 |
> -- |
657 |
> gentoo-performance@g.o mailing list |
658 |
> |
659 |
> |
660 |
> -- |
661 |
> gentoo-performance@g.o mailing list |
662 |
> |
663 |
> |
664 |
> -- |
665 |
> gentoo-performance@g.o mailing list |
666 |
> |
667 |
> unsubscribe |
668 |
> |
669 |
> |
670 |
> |
671 |
> Okay guys, stop mailing "unsubscribe" to the list. Instead, send a |
672 |
> blank message to gentoo-performance+unsubscribe@l.g.o |
673 |
> |
674 |
> alright? |
675 |
> -- |
676 |
> gentoo-performance@g.o mailing list |
677 |
> |
678 |
> -----BEGIN PGP SIGNED MESSAGE----- |
679 |
> Hash: SHA1 |
680 |
> |
681 |
> lnxg33k wrote: |
682 |
> > Thanks Alec Warner for the great explanation. It still seems like by not |
683 |
> > having portions of the tree by using EXCLUDEFORM and deleting the local |
684 |
> > dirs that you'd save some time in the --metadata part of sync as less |
685 |
> > ebuilds are available to be checked. Is this simply a wrong |
686 |
> misconception? |
687 |
> |
688 |
> The portion that "updates metadata cache" has nothing to do with what |
689 |
> ebuilds are in the tree. It simply takes the server-side caches ( |
690 |
> pregenerated ) and sync them into your local cache. You could RSYNC |
691 |
> exclude the whole tree and this will still happen for every package, |
692 |
> since the metadata is generated on the server ( and the server has the |
693 |
> whole tree ). |
694 |
> |
695 |
> Now if you were to rsync exclude metadata/ you would reduce the |
696 |
> - --metadata portion...because it wouldn't happen ;) |
697 |
> |
698 |
> - -Alec Warner (antarus) |
699 |
> -----BEGIN PGP SIGNATURE----- |
700 |
> Version: GnuPG v1.4.2 (GNU/Linux) |
701 |
> Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org |
702 |
> |
703 |
> iQIVAwUBQ9GCJ2zglR5RwbyYAQLajw/+Mvfu9+0Sfo8nHo7gbNytHRNL1jVI61RD |
704 |
> /EIiZZwV067X77IP72o0Y7SICRvnooEqtLvQW+rd2c/Kb6W4EtDga94X8EbOrHIm |
705 |
> 0/IFqg7OqUa/psU6IkaPk959u7UJTnqlWmzluLbqRTx/X03lpjk6n+V4BOTRhcTA |
706 |
> WHa80quSpd5EkfdFAJ1oWVMn9ck7xSTn3ulzmlznCkLdWK6iR+m+r+wAWMPlcRFG |
707 |
> xE/Ik5uMVemlWuAElhMoB4xr/2hKfcsu/Egw9F+zVL5+3mGMyygHhELAvTVHU/C9 |
708 |
> jLX3noNFC+xSOesmC0e+l88Uyz2AYPuhg8S+3qciC+4Ax2QkDhhRxwM7lLF6yPBx |
709 |
> UpDNnTecT1iVuFUhHP08T2GPq8Nyvtzj4oqgjzKq1+l2RdehDM8j0KZrq5ZwDszb |
710 |
> CdKakVUrVXmMOEp16E48k5/66sulgJ5fNdVubJYNdPwsY2dNJ5MC7wbxSwIT46TD |
711 |
> 88tYAgco4cH9o4whwL2KZIjCQodKgvNwCnvW3qeXOdD/WqRSvuVbFIZh/+YZMHXi |
712 |
> KVnWrePS2kwukMiL28oB9fwsJomQpwxCJlhZxuLto7kM1vBhlKLOeFfoZ8gm6mrF |
713 |
> gBkDwiyV7aNHL4tFgAGtvDPReQhRS4DcN61EVEOYxMxQIKh1lDSFR1UW+j538S9c |
714 |
> J/7XGJI9CxQ= |
715 |
> =Qike |
716 |
> -----END PGP SIGNATURE----- |
717 |
> -- |
718 |
> gentoo-performance@g.o mailing list |
719 |
> |
720 |
> Alec Warner wrote: |
721 |
> > -----BEGIN PGP SIGNED MESSAGE----- |
722 |
> > Hash: SHA1 |
723 |
> > |
724 |
> > lnxg33k wrote: |
725 |
> >> Thanks Alec Warner for the great explanation. It still seems like by |
726 |
> not |
727 |
> >> having portions of the tree by using EXCLUDEFORM and deleting the local |
728 |
> >> dirs that you'd save some time in the --metadata part of sync as less |
729 |
> >> ebuilds are available to be checked. Is this simply a wrong |
730 |
> misconception? |
731 |
> > |
732 |
> > The portion that "updates metadata cache" has nothing to do with what |
733 |
> > ebuilds are in the tree. It simply takes the server-side caches ( |
734 |
> > pregenerated ) and sync them into your local cache. You could RSYNC |
735 |
> > exclude the whole tree and this will still happen for every package, |
736 |
> > since the metadata is generated on the server ( and the server has the |
737 |
> > whole tree ). |
738 |
> > |
739 |
> > Now if you were to rsync exclude metadata/ you would reduce the |
740 |
> > - --metadata portion...because it wouldn't happen ;) |
741 |
> > |
742 |
> <snip> |
743 |
> |
744 |
> Ah, ok. Hearing it again makes it sound more reasonable. I'll have to |
745 |
> watch my |
746 |
> rsync output a bit more closely. Thanks again for the explanation. |
747 |
> -- |
748 |
> gentoo-performance@g.o mailing list |
749 |
> |
750 |
> How to get Maximum performance in Graphics on Nvidia Drivers???? |
751 |
> |
752 |
> |
753 |
> -- |
754 |
> gentoo-performance@g.o mailing list |
755 |
> |
756 |
> |
757 |
> |