Database Users

[Q]database read/write speed

Hi all,
I was doing some test with a friend of mine that have a galaxy s with ext4 lag fix.
We have noticed that in quadrant there is a huge different in time during the passages "writing database" and "reading database".
So is the desire hd filesisten not well optimized (not like the original galaxy s one ) and is there a way to speed up read/write operations?
Thank u all
Sent from my Desire HD using XDA App

by default the galaxy S out of the box uses RFS which when doing IO to DB can be very very slow....because it does a lot of journalling.
I believe ext4 does not do journalling?? correct me if i am wrong...
if your friend uses the OCLF then ...believe me over a period of 1 week..or after your friend has put more than 10 apps...it will start to degrade...
RFS in the GS corrupts the FS over time...
I have had my DHD for like 2 weeks and installed over 50 apps..
quadrant scores are always...close to 2000..
using Froyo on the GS without any lag fix u get approx 1000 quadrant score..
with lag fix you get close to 2000..but then again is temporary..
if your friend had applied the voodoo lag fix..then i thing that uses a different journalling policy.and on Froyo GS gets approx around the 1500...
so DHD is indeed optimize IMO

I came from the galaxy s. Wich i loved, after much work and tweaking. Straight out of the box it was useless. But back to topic. Of course it would be possible but i dont really see the need. Itryed all the diffrent lagfixes for SGS, OCLF wich uses EXT2, Voodoo EXT4, and Ztupys kernel with ddiferent options and on Android 2.1 they aall really boosted the Quadrant scores was getting 2200+. But on Android 2.2 it didnt have same effect quadrant about 1500-1700 still good fix. But when i run Quadrant on DHD, in I/O section it stops a little while r/w data but it did that with lagfix on SGS also. Only lagfix that made it run thru with out pauseing was OCLF wich uses EXT2, and Ext2 gets corrupted or something after a while and performance slows down.
I think we have a good stable device and i dont see the need.

nandihno said:
I believe ext4 does not do journalling?? correct me if i am wrong...
Click to expand...
Click to collapse
I am afraid not, ext4 is basically ext3 with journalling c.f Wikipedia

thank u all for replies. Mine was just curiosity more than need because DHD in every day use is really fast and responsive (not like my old iphone 3g....)
I think i has applied the OCLF so it will be degraded in some times, and he said also that he reboot the phone once a day..
Sorry to mod if i posted in the wrong section
muffy

"lagfix" does nothing to Epic?

I've always been curious about other file systems rather than rfs on OneNAND devices and their performances.
For example, supercurio, a famous developer of voodoo lagfix/color/sound, formatted /dbdata on i9000 Galaxy S (which is also OneNAND) and reports better performance, although I can't find any evidence.
Since rfs is basically FAT with POSIX and some other stuffs, other "advanced" file systems will do better performances-- this is what I thought. Yes there is no lag that occurs on other Galaxy S on Epic 4G, and it's because our OneNAND is physically superfast NAND flash with SLC NAND, NOR logic and SRAM buffer. Much faster than moviNAND(of Galaxy S) or other NAND flash. That's all. That doesn't mean changing file system from rfs to another will do nothing. With rfs the performance is good enough already. The question is does other file systems increase the performance.
(and Yes quadrant is a terrible benchmark software. It does not give any criterion about its benchmark, and boosts too much points on I/O.)
Deceived by Quantum Rom's quadrant tmpfs trick, I decided to install VIPERrom's ext2 "lagfix" (although I don't want to call it "lag fix" because there is no lag on our Epic 4G) and run RL benchmark of SQLite performance. Since SQLite is sometimes quite relevant to real usage in our Android life, especially for database-using applications, please don't call it mere benchmark. Anyway, the result gives:
In order of
1000 INSERTs /
25000 INSERTs in a transaction /
25000 INSERTs into an indexed table in a transaction /
100 SELECTs without an index /
100 SELECTs on a string comparison /
Creating an index /
5000 SELECTs with an index /
1000 UPDATEs without an index /
25000 UPDATEs with an index /
INSERTs from a SELECT /
DELETE without an index /
DELETE with an index /
DROP TABLE /
Overall
Normal, rfs, DK05 FroYo:
11.738 / 2.91 / 2.903 / 0.096 / 0.076 / 1.095 / 2.442 / 6.618 / 6.908 / 2.524 /
2.991 / 2.651 / 1.998 / Overall 44.95
VIPER ext2 patched, DK17 VIPERrom FroYo:
0.971 / 2.758 / 2.745 / 0.091 / 0.073 / 0.989 / 2.488 / 6.749 / 6.951 / 1.68 /
1.21 / 1.227 / 0.308 / Overall 28.24
(For comparison:
Normal, rfs, Korean M110S Galaxy S, SK22 FroYo (moviNAND)
81.281 / 5.857 / 4.107 / 0.251 / 0.117 / 1.221 / 2.26 / 7.181 / 13.065 / 4.359 / 7.754 / 10.481 / 11.135 / Overall 149.069)
The answer is obvious: ext2 increases I/O performance.
The new questions are:
1) is ext3 or ext4 better than ext2?
2) is it safe to use ext2 on OneNAND devices?
(there are several debates about ext's on moviNAND as well.)

Forgot to mention: it is not possible to support ext3 or 4 at the moment for FroYo now, since there is no source code released yet. (which is of course thing; there is no official froyo yet lol)

chocoberry said:
Forgot to mention: it is not possible to support ext3 or 4 at the moment for FroYo now, since there is no source code released yet. (which is of course thing; there is no official froyo yet lol)
Click to expand...
Click to collapse
If you had been following the Quantum Rom thread, you would have seen that the "Lagfix" was nothing more than a joke to prove a point about Quadrant, and other epic roms using a lagfix.
Froyo is not complete yet, there is no source, there is still debugging code, and it is not optimized for release. This will explain all low scores. Also; issuing a ton of sqlite queries hardly amounts to real life performance on a phone, and once again it is all I/O. (As sqlite is stored in files.) (And yes, I am aware that we use sqlite to store data. But we don't do thousands of transactions in an operation... ever) The same thing happens when you loop an ext FS over RFS, you get inflated values because of the ram buffer between RAM and ext filesystem.
When froyo goes official; I will try my hand at porting a ext4 fs to the epic (Minus journaling, as it kills performance); but Samsung is making improvements to rfs, and the results are getting better. However, you need to realize that I/O performance is not all that matters on a phone, and regardless there are other aspects that will throttle you anyway.

Dameon87 said:
you need to realize that I/O performance is not all that matters on a phone, and regardless there are other aspects that will throttle you anyway.
Click to expand...
Click to collapse
Yes you are true. For epic it's not the I/O which makes bottleneck (not like other SGS), obviously, because it is already on top of all Android devices even with rfs, thanks to OneNAND flash.
But the purpose of thread and benchmark is to test "if" the file system change will increase I/O performance or not on OneNAND devices (which now runs on Sammy's propriety driver). And my conclusion is that it "does" increase performance. That's all. I know that you are a ROM developer (and I'm a Quantum ROM user also) who hears tons of useless quadrant things, but you don't have to do that to me. You overinterpreted the data; I don't want to emphasize or imply anything.

chocoberry said:
Yes you are true. For epic it's not the I/O which makes bottleneck (not like other SGS), obviously, because it is already on top of all Android devices even with rfs, thanks to OneNAND flash.
But the purpose of thread and benchmark is to test "if" the file system change will increase I/O performance or not on OneNAND devices (which now runs on Sammy's propriety driver). And my conclusion is that it "does" increase performance. That's all. I know that you are a ROM developer (and I'm a Quantum ROM user also) who hears tons of useless quadrant things, but you don't have to do that to me. You overinterpreted the data; I don't want to emphasize or imply anything.
Click to expand...
Click to collapse
Oh no. I wasn't trying to come off as an ass at all. I was just giving some information.
We will not know any concrete evidence until official 2.2 is released as sammy has been working on rfs.
Sent from my SPH-D700 using Tapatalk

As I said before..the lagfix would make the phone into a nice database server lol
But part of deploying a database has always been choosing the right file system...:/
Its not like you won't get any increase at all in real life performance...I'd probably guess at around 1% or so..but obviously not enough to justify what people perceive Quadrant shows...hence if your not using Quadrant Pro..really no point of using Quadrant much :/

Dameon87 said:
When froyo goes official; I will try my hand at porting a ext4 fs to the epic (Minus journaling, as it kills performance); but Samsung is making improvements to rfs, and the results are getting better. However, you need to realize that I/O performance is not all that matters on a phone, and regardless there are other aspects that will throttle you anyway.
Click to expand...
Click to collapse
isnt ext2 the fast one, ext 3 the data secure one, and ext4 the best of both worlds? when you removing journaling isnt that basically going to just drop it down to ext2?

[Q] For best performance on swiftdroid M5!

Hi guys!
I just wanted to ask you about what we Gt540 users should do for have best performance on swiftdroid M5.
Like in Quadrant and other benchmark programs.

DarkCaine said:
Hi guys!
I just wanted to ask you about what we Gt540 users should do for have best performance on swiftdroid M5.
Like in Quadrant and other benchmark programs.
Click to expand...
Click to collapse
kill processes, overclock, maybe make SWAP...

Actually, without any change, on my oppinion, it's the best ROM with best performance.
I set clock on max 768/min 245; swap partition with swappines 10, s2e (400 MB ext3 partition) and it's working better then ever!
Best of all battery is on 60%, and it's already 2d1h on battery!
Nice work Mur4ik!

max1919 said:
kill processes, overclock, maybe make SWAP...
Click to expand...
Click to collapse
i think...Swap is the most important

maryyo said:
Actually, without any change, on my oppinion, it's the best ROM with best performance.
I set clock on max 768/min 245; swap partition with swappines 10, s2e (400 MB ext3 partition) and it's working better then ever!
Best of all battery is on 60%, and it's already 2d1h on battery!
Nice work Mur4ik!
Click to expand...
Click to collapse
That sounds cool but what is s2e? Am gonna try imitating you. xD

nibras_reeza said:
That sounds cool but what is s2e? Am gonna try imitating you. xD
Click to expand...
Click to collapse
simple2ext...it's an application for moving applications and cache (internal memory) on ext partition. You have to create ext (i have ext3) partition and download app from market.
And if you really wont to copy me then you need swap partition as well.
I was generous so i set 512 MB ext3, and 256 swap, but you don't need so much (256ext/64swap is enough i think!
...if you put all that, let me now on your oppinion...

I have a 512MB ext2. I have already used 384MB on ext2. It includes data(lib), dalvick cache and apks. I use link2sd. I will check out s2e.
Does it move app cache or the cache partition?

nibras_reeza said:
I have a 512MB ext2. I have already used 384MB on ext2. It includes data(lib), dalvick cache and apks. I use link2sd. I will check out s2e.
Does it move app cache or the cache partition?
Click to expand...
Click to collapse
I think app cache (if this is app folder), but not the cache partition!
...so on ext partition are folders: app, app-private, dalvik-cache, download, lost+found.

yeah...just overclocking and SWAP, also get an app called ROOT TOOLS.
it allows you to change boot animation, over clock, ad block(works well with angry birds), properties builder, and some other tweaks and hacks.
and btw can anyone link me a copy of livewallpaper picker apk form the original m5 rom? im using data2system

I use AdFree for Ad blocking on Android and I use MVPS Hosts file on my PC. I prefer HOSTS based ad blocking more than other methods.
And, hmm, did you guys check LBE Privacy Guard? It's pretty cool.
And, as for app cache, I guess Link2SD would do a better job since you can individually select which app(cache,apk,lib) goes to SD and which stays internal. This way, you could have Dialer, Keyboard and essential apps always on phone even if the SD is removed or not loaded right after boot.

I spend whole afternoon yesterday to make Link2SD and later S2E to work, didn't manage it , and now I move apps in .android_secure. Oh well...
My Quadrant scores:
436 - stock 2.1 ROM
800 - Swiftdroid v2.0 M5
1017 - Swiftdroid v2.0 M5 (OC @ 806 MHz)
I installed SD-Booster. And that's the great way to boost performances of microSD card .

maur4ik said in v2.0 m5 no need to overclock your phone because it has a built in overclock and it is a fastest either for some previous version of swiftdroid

angemon10 said:
I spend whole afternoon yesterday to make Link2SD and later S2E to work, didn't manage it , and now I move apps in .android_secure. Oh well...
My Quadrant scores:
436 - stock 2.1 ROM
800 - Swiftdroid v2.0 M5
1017 - Swiftdroid v2.0 M5 (OC @ 806 MHz)
I installed SD-Booster. And that's the great way to boost performances of microSD card .
Click to expand...
Click to collapse
About S2E, i had a lot of problems too. I tried to create ext in recovery more then a million times, and i finally make S2E to work when i created ext3 on computer with Acronis Disk Director.
About SD-Booster, i tried it as well as one or two more similar apps and i didn't notice any difference in quadrant nor in Antutu.

maryyo said:
About S2E, i had a lot of problems too. I tried to create ext in recovery more then a million times, and i finally make S2E to work when i created ext3 on computer with Acronis Disk Director.
About SD-Booster, i tried it as well as one or two more similar apps and i didn't notice any difference in quadrant nor in Antutu.
Click to expand...
Click to collapse
Quadrant does not check the SD speeds.
The performance boost of SD booster should be checked with SD Tools ( https://market.android.com/details?id=ales.veluscek.sdtools ) or something similar.
And, quadrant score is not always the best measure of a phone's performance. A better alternative would be to check the FPS of a graphics heavy application.
Overclocking speeds up only the CPU. But, in most mobile devices, CPU is not the usual limiting factor of performance. Overclocking will show a performance hike only in arithmetic apps. Rather, on most phones, performance is limited by IO speeds of the hardware.

nibras_reeza said:
Quadrant does not check the SD speeds.
The performance boost of SD booster should be checked with SD Tools ( https://market.android.com/details?id=ales.veluscek.sdtools ) or something similar.
And, quadrant score is not always the best measure of a phone's performance. A better alternative would be to check the FPS of a graphics heavy application.
Overclocking speeds up only the CPU. But, in most mobile devices, CPU is not the usual limiting factor of performance. Overclocking will show a performance hike only in arithmetic apps. Rather, on most phones, performance is limited by IO speeds of the hardware.
Click to expand...
Click to collapse
Correct, not in quadrand then in Antutu (it measure SD), and i checked speed in SD Tools to, and the results were the same.
When you mentioned it, haw can I check the FPS? (And, I think quadrant does the same thing. It measure FPS in few 2D and 3D simulations)
My Quadrant score: 1174 - Swiftdroid v2.0 M5 (OC @ 768 MHz)
P.S. I just flashed M6...I'm very curious...

angemon10 said:
I spend whole afternoon yesterday to make Link2SD and later S2E to work, didn't manage it , and now I move apps in .android_secure. Oh well...
My Quadrant scores:
436 - stock 2.1 ROM
800 - Swiftdroid v2.0 M5
1017 - Swiftdroid v2.0 M5 (OC @ 806 MHz)
I installed SD-Booster. And that's the great way to boost performances of microSD card .
Click to expand...
Click to collapse
lol I get 1250+ even when on v1.8 ! OC @ 825mhz
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
In this Im using [email protected]

maryyo said:
Correct, not in quadrand then in Antutu (it measure SD), and i checked speed in SD Tools to, and the results were the same.
When you mentioned it, haw can I check the FPS? (And, I think quadrant does the same thing. It measure FPS in few 2D and 3D simulations)
My Quadrant score: 1174 - Swiftdroid v2.0 M5 (OC @ 768 MHz)
P.S. I just flashed M6...I'm very curious...
Click to expand...
Click to collapse
I got some performance boost with SD Boost but it was just some 2MB/s or so which was insignificant since it can decrease SD life.
There are a lot of graphics benchmarking apps.
You can try neocore or An3DBench. There was another OpenGL test app in market under demo section but I can't remember name. I haven't personally tried many of these but this is just a start. You can look for "3d benchmark android" on Google. You can also check the FPS in many graphics intensive games.

V6 SuperCharger + SD Booster (set on 2048) + 128 MB Linux swap partition =

angemon10 said:
V6 SuperCharger + SD Booster (set on 2048) + 128 MB Linux swap partition =
Click to expand...
Click to collapse
If we talking about best performance, V6 Supercharger is definitely must have!
I set it yesrerday (so, it's still on probation), and it's like i'm having a new phone!
Sent from my GT540 using XDA Premium App

maryyo said:
If we talking about best performance, V6 Supercharger is definitely must have!
I set it yesrerday (so, it's still on probation), and it's like i'm having a new phone!
Sent from my GT540 using XDA Premium App
Click to expand...
Click to collapse
I use Auto Memory Optimizer to modify memory management settings. Easier than tinkering with scripts.

Best Lagfix for SGS Android 2.3.3/2.3.4

What is Best Lagfix for SGS Android 2.3.3/2.3.4

I dont think you need lagfix anymore on 2.3.3 and 2.3.4. It is already fast, smooth and stable. Though there are some bugs that are not related to lags

Modern lagfixes are mostly just kernels supporting an ext-4 conversion - they all will do more or less the same job. Pick one you like the features of, install and convert.
However as has been mentioned lag has been reduced a lot without it so the improvement is a lot less noticeable than it used to be.

since the GB roms there is no real need for "lagfix"
If you want you can still convert to ext4, just flash any kernel supporting it

Actually there is no noticeable difference between ext4 and rfs fs on gb roms in terms of UI smoothness and speed. And yeah I am using ext4 right now so that is just my personal observation.

jbdroid said:
Actually there is no noticeable difference between ext4 and rfs fs on gb roms in terms of UI smoothness and speed. And yeah I am using ext4 right now so that is just my personal observation.
Click to expand...
Click to collapse
I second this. Lately I've just been using RFS, and I feel it's just as fast as when I ran ext4.

EXT 4 Voodoo lagfix

ext4 rules!
I think ext4 conversion still matters.
I use feedR to follow my google reader feeds. With rfs I can't open posts smoothly and fast as it is with ext4. Actually with rfs it's annoyingly laggy that I can't catch up with all the posts in the feeds.

I'm using feedr to follow 15 websites with rfs, and there is no problem with smoothness or feeds.
Sent from HAL-I9000 using Tapatalk

between rfs n ext4, in GB , yes they r not laggy.
but , how about battery consumption? rfs or ext4 that use alot of battery?

syamsoul said:
between rfs n ext4, in GB , yes they r not laggy.
but , how about battery consumption? rfs or ext4 that use alot of battery?
Click to expand...
Click to collapse
I didnt notice any difference in battery consumption. RFS or EXT4, i usually get 2-3 days of mild usage, 1-2 days of heavy usage, on both.

K0v4L said:
I'm using feedr to follow 15 websites with rfs, and there is no problem with smoothness or feeds.
Sent from HAL-I9000 using Tapatalk
Click to expand...
Click to collapse
Try opening post one after other immediately.

i've been on ext4 ever since eclair, never complained. I keep it like this

Try voodoo lagfix,but on gingerbread i think you don't have to use any lagfix.
Sent from my GT-I9000 using XDA App

Well, my JVH lags as hell...

I don't think you will notice any real difference with different lagfixes. I tried most types and didn't even notice real difference with rfs.
Most devs think ext4 is better so I do apply a lagfix, but not really sure why (better quadrant score? doesn't say anything but still)

I used couple of ROMs from XDA (F1, F1 SGS2, Ficeto, MIUI..) without lagfix... I am using phone a lot (battery lasts max 15-20hrs)... So, There is small diferences between them... I used them without lagfixes... Now, I installed JVP 2.3.4 Stock, Galaxian Kernel, and Converted to EXT4...
Personal opinion:
1. STOCK 2.3.4 JPV - Not bad, Fast, Works fine
2. +Galaxian Kernel - Faster 20-25% at least (using CPU Master because of OC 1.4GHz)
3. EXT4 Convert -> This speeded up my system by 20-30% more...
Explanation.. RFS file system is better with small files (under 5KBytes), and EXT4 is faster and a lot better with larger files... So, when you work with pictures, videos, and applications that uses bigger files then text, EXT4 shows its best side...
Now, my device is focused on Bigger apps, Pictures and HD Videos, so EXT4 is much faster for me...
Use FS that is better for your use...
P.S. EXT4 has faster R/W rate, so when recording, copying and similar, you will see difference...

can't wait until someone gets ubifs file system working, by all accounts should be awesome
Sent using geek power

Am on gb but i use ext4 lagfix because after number of apps increases, it lags!!

I've tried out 2.2 with lagfix and 2.3 without lagfix, but I find that now 2.3 is slower than 2.2 with lagfix. Not sure why this is true, but I noticed that starting apps was slower.

I/O Scheduler - A quick test

After trying to decide which I/O scheduler to use, I decided to try running the Androbench storage performance benchmark tool for the I/O schedulers,
cfq, deadline, noop,vr and sio to find out.
All tests run on deodexed JVQ on Galaxian [email protected]
The file system used was /data (ext4)
I used standard buffer sizes (256Kb for sequential I/O) and (4kb for random I/O) To speed the tests up, I used a read file size of 16Mb and a write file size of 1Mb
I did test using the recommended sizes of 32MB and 2MB) but the rates were the same so I kept the smaller sizes to speed up the tests
I know its not very scientific but provides a rough estimate of the relative performance.
MBRS =MB/s sequential reads
MBWS =MB/s sequential writes
RRIOPS =random reads IO's/sec
RWIOPS=random writes IO's/sec
Test #1
MBRS MBWS RRIOPS RW IOPS
cfq 19.40 4.8 1014 35
deadline 19.39 6.1 1119 42
noop 19.47 9.0 1098 43
vr 17.76 7.7 1105 45
sio 18.30 6.9 1152 49
Test #2
MBRS MBWS RRIOPS RWIOPS
cfq 19.18 3.3 1057 33
deadline 26.27 5.5 1151 43
noop 19.08 8.6 1059 41
vr 19.55 6.4 1120 50
sio 19.16 6.3 1122 45
Test #3
MBRS MBWS RRIOPS RWIOPS
cfq 19.25 3.8 1120 34
deadline 18.97 7.8 1146 43
noop 19.59 9.3 1161 45
vr 19.14 6.6 1227 53
sio 19.46 7.5 1172 50
my conclusion - use NOOP or SIO(if stable)
As you can see, sequential reads (MB/s) and sequential IOPS are very high so I don't think we need to worry about these too much, as you know (from Quadrant) it's writes that are VERY slow
Unsurprisingly, we can see that the slowest scheduler is the default CFQ.
The noop scheduler produced the fastest sequential write rate on all runs.
the VR and SIO schedulers produce the greatest number of IOPS for random writes.
Think I will use the noop scheduler as it uses little CPU and is fairly well established and its not far off the random write IOPS rate.
Is someone can be arsed, I think we should test with all background processes off, using a 1Ghz speed, and run say 10 sets of tests and order the results using a spreadsheet.
Would be good to see these results run against a RFS file system too!!!
Would also be good to test I/O against say a /cache file system with journalling switched off to unserstand the actual improvement.
I am surprised at the read speed, the IO rate is equivalent to 5 or 6 15K SA/FC physical disks - Whahoo!!!!!

Hi,
Last evening I updated my kernel to Galaxian 2.0 and after reading your test results I set my I/O scheduler to noop and this morning... I found my phone hot, checked the battery status and had the nasty surprise of finding that Android OS has taken 41% of my battery and the phone didn't sleep almost all night. Did you have this problem? I'm trying to figure out whether the Android OS bug returned because of the kernel update or the I/O scheduler change.
Thanks.

Hmm, used system panel to check android and overhead for me was very low, I use juice defender and lose about 2 or 3% battery overnight, phone is cool, try wiping cache, dalvik and fix permissions and reboot once or twice, if the same change to deadline and observe difference
Sent using geek power

I am getting quite different values
Seq read 12,75 MB/s
Seq write 4.0 MB/s
Random read 1596 IOPS(4k), 6,23MB/s
Random write 41,74 IOPS(4k), 0,16 MB/s

Brotuck said:
I am getting quite different values
Seq read 12,75 MB/s
Seq write 4.0 MB/s
Random read 1596 IOPS(4k), 6,23MB/s
Random write 41,74 IOPS(4k), 0,16 MB/s
Click to expand...
Click to collapse
which scheduler you got that results?
Im going to test noop with semaphore 1.3.3

My benchmark results!
Hey, I read countless threads on I/O schedulers, but most of them were based on opinions or feelings... except this one! I want to contribute, so I decided to run my set of benchmarks too.
I used the latest AndroBench 2.0 from the market, default/recommended settings. Each I/O scheduler was tested three times, and the numbers below are the averages of the results.
Samsung Galaxy S GT-I9000
Firmware: stock 2.3.3 w/root (RSJV3, latest official firmware in Italy)
Kernel: Galaxian 2.4
File system: RFS
No overclock
Cpu governor set to "performance" to make sure frequency doesn't fluctuate during test runs.
I created a couple of simple comparison indexes:
Read Index = SeqRead * RndReadMB * RndReadIOPS
Write Index = SeqWrite * RndWriteMB * RndWriteIOPS * 100
Here are the averages. You can download the spreadsheet with all the results, it's attached to my post.
--------------noop-----vr-------sio------cfq------deadline
Read Index....57750....57305....54328....58849....60013
Write Index...31339....24051....27602....23233....18628
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
My conclusion: I agree with gsw5700, NOOP is definetely the most interesting I/O scheduler for our SGS even on RFS file system. Satisfactory read speed and best-of-the-class write speed!

Here are my results (average of 3 runs)
seq.read seq.write r-read r-write
VR: 12.5 - 4.22 - 5.82 - 0.21
SIO: 12.61 - 3.86 - 6,07 - 0.207
BFQ: 12.8 - 3.93 - 6,01 - 0,18
Noop:12.6 - 4,05 - 6,03 - 0,208
(seq= sequential; r-= random; all values in MB/s)

I've heard BFS is supposedly very good, but can't find any data. Could someone run a test of it against Noop etc. so we can all see...
Thanks
Mike
Garbled meaning induced by swype when posting from XDA app on SGS I9000.

Uhm... never heard of BFS before, which kernel do you have?

ZioGTS said:
Uhm... never heard of BFS before, which kernel do you have?
Click to expand...
Click to collapse
BFS is most likely brain **** scheduler.

TalonDEV 0.4.2 0.4.3
added BFQv2-r1 I/O Scheduler (SIO default)

BFS is a CPU scheduler which has nothing to do with Noop,sio,vr,cfq which are i/o schedulers
A BFQ test is a few posts up....

.:Crack:. said:
Here are my results (average of 3 runs)
seq.read seq.write r-read r-write
VR: 12.5 - 4.22 - 5.82 - 0.21
SIO: 12.61 - 3.86 - 6,07 - 0.207
BFQ: 12.8 - 3.93 - 6,01 - 0,18
Noop:12.6 - 4,05 - 6,03 - 0,208
(seq= sequential; r-= random; all values in MB/s)
Click to expand...
Click to collapse
I did another test with BFQ and noop (5 runs each) and noop won.
(BTW i was a V(R) fanboy before )

.:Crack:. said:
BFS is a CPU scheduler which has nothing to do with Noop,sio,vr,cfq which are i/o schedulers
A BFQ test is a few posts up....
Click to expand...
Click to collapse
Whoops! Sometimes a little bit of knowledge makes me dangerous!... But this time I got away with just looking silly!!!
Thanks for correcting me... making mistakes its often the best way to learn!
Mike
Garbled meaning induced by swype when posting from XDA app on SGS I9000.

Interesting test, however I am troubled with my results. I tried running this Androbench software and it seems that I have a lot of fluctuations.
For example during one test I get 4.92mb/s write speed @ noop and then during second test I get 1.52mb/s write speed @ noop test and last one was 2.02mb/s. I mean write speed is write speed you can have minor fluctuations say 4.92mb/s on the first run and 4.90 on the second run. This is of course while the device was not used.
If I run a benchmark software on my pc, I will get almost all the time identical results.
So either there is something wrong with teh software or something else is at play.
Also this guy from over here: http://www.team-continuum.net/home/index.php?topic=290.0 claims that operating on different CPU governors also has impact on benchmark results.

Theoretically, SIO is the best.
Benchmark can't represent real life situations as I/O transactions are mostly spurious rather than 1 big chunk of data to and fro (on which this is what benchmarks are doing).
More info at my sig (Android Optimization Tips). CTRL-F for scheduler.

What does the phone do most of, i guess it reads much more than it writes so wouldn't it be better to use a fast reading with satisfactory write speeds scheduler?

bobdoblo said:
What does the phone do most of, i guess it reads much more than it writes so wouldn't it be better to use a fast reading with satisfactory write speeds scheduler?
Click to expand...
Click to collapse
+1
and what about the row sceduler
row supposed to be the best fro read, its designed for read or im mistaken?

herc2k said:
+1
and what about the row sceduler
row supposed to be the best fro read, its designed for read or im mistaken?
Click to expand...
Click to collapse
...I ran some tests for schedulers, including row and for the test the row seems to be the best, I'm using it on JB 4.2...

What about the Zen and Row I/O schedulers?