<div class="snippet-clipboard-content notranslate position-relative overflow-auto" data-snippet-clip

Here is the output if I reset everything: <div class="snippet-clipboard-content no

Support for NVMe devices / dm-crypt / LUKS / LVM about s HOT 10 CLOSED

teohhanhui commented on August 21, 2024

Support for NVMe devices / dm-crypt / LUKS / LVM

from s.

Comments (10)

Algodev-github commented on August 21, 2024

Hi, sorry for the late reply. I'm not sure your configuration works well (it may). The output of the script may probably help me understand it. At any rate, the right way to tell S which partition you want your test performed on, is to set TESTD_DEV, in ~/.S-config, to that partition. In your case, I guess you should just set TEST_DEV=nvme0n1p2 For your convenience, I'm pasting the full documentation on that parameter, as you find it in ~/.S-config. Before pasting that, let me tell you that any extension to fully support dm would be warmly welcome! :) Thank you for trying this suite, Paolo # Set the following parameter to the name or the full path of a device # or of a partition, if you want to perform tests on that device or # partition. In the next comments, we assume for simplicity that # TEST_DEV is set to just the name of a device. # # The following two alternatives are handled differently: first, the # device contains the partition ${TEST_DEV}1 (or ${TEST_DEV}p1, # depending on the type of device) and ${TEST_DEV}1 contains a mounted # filesystem; second, the latter compound condition does not hold. In # the first case, all test files are created in that filesystem. In # the second case, the execution is simply aborted if the following # FORMAT parameter is not set to yes. If, instead, FORMAT is set to # yes, then # - the device is formatted so as to contain a partition ${TEST_DEV}1 # or ${TEST_DEV}p1; # - an ext4 filesystem is made on that partition; # - that filesystem is mounted. # # If you go directly for a partition, then the latter must contain a # mounted filesystem. # # In all succesful cases, all test files are created in the filesystem # contained in the test partition. # # If TEST_DEV is set, then BASE_DIR will be built automatically, # overriding your possible choice below. # # Be careful in setting TEST_DEV manually in case of bcache or # raids. With these configurations, more then one device is involved. # The simplest option is to set BASE_DIR to a directory stored on # these devices, and let the code of the suite automatically detect # all the involved devices. TEST_DEV=

…

Il giorno 26 feb 2020, alle ore 16:09, Teoh Han Hui ***@***.***> ha scritto: $ df -h Filesystem Size Used Avail Use% Mounted on dev 7.7G 0 7.7G 0% /dev run 7.8G 2.1M 7.8G 1% /run /dev/dm-0 468G 220G 225G 50% / tmpfs 7.8G 209M 7.6G 3% /dev/shm tmpfs 7.8G 0 7.8G 0% /sys/fs/cgroup tmpfs 7.8G 46M 7.7G 1% /tmp /dev/nvme0n1p1 599M 48M 552M 8% /boot/efi /dev/loop0 55M 55M 0 100% /var/lib/snapd/snap/core18/1668 /dev/loop1 161M 161M 0 100% /var/lib/snapd/snap/gnome-3-28-1804/116 /dev/loop2 157M 157M 0 100% /var/lib/snapd/snap/gnome-3-28-1804/110 /dev/loop4 55M 55M 0 100% /var/lib/snapd/snap/core18/1288 /dev/loop3 92M 92M 0 100% /var/lib/snapd/snap/core/8592 /dev/loop5 256K 256K 0 100% /var/lib/snapd/snap/gtk2-common-themes/9 /dev/loop6 92M 92M 0 100% /var/lib/snapd/snap/core/8689 /dev/loop7 45M 45M 0 100% /var/lib/snapd/snap/gtk-common-themes/1353 /dev/loop8 45M 45M 0 100% /var/lib/snapd/snap/gtk-common-themes/1440 /dev/loop9 256K 256K 0 100% /var/lib/snapd/snap/gtk2-common-themes/5 tmpfs 1.6G 956K 1.6G 1% /run/user/1000 $ lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT loop0 7:0 0 54.7M 1 loop /var/lib/snapd/snap/core18/1668 loop1 7:1 0 160.2M 1 loop /var/lib/snapd/snap/gnome-3-28-1804/116 loop2 7:2 0 156.7M 1 loop /var/lib/snapd/snap/gnome-3-28-1804/110 loop3 7:3 0 91.3M 1 loop /var/lib/snapd/snap/core/8592 loop4 7:4 0 54.6M 1 loop /var/lib/snapd/snap/core18/1288 loop5 7:5 0 132K 1 loop /var/lib/snapd/snap/gtk2-common-themes/9 loop6 7:6 0 91.4M 1 loop /var/lib/snapd/snap/core/8689 loop7 7:7 0 44.2M 1 loop /var/lib/snapd/snap/gtk-common-themes/135 loop8 7:8 0 44.9M 1 loop /var/lib/snapd/snap/gtk-common-themes/144 loop9 7:9 0 132K 1 loop /var/lib/snapd/snap/gtk2-common-themes/5 nvme0n1 259:0 0 477G 0 disk ├─nvme0n1p1 259:1 0 600M 0 part /boot/efi └─nvme0n1p2 259:2 0 476.4G 0 part └─luks-d57f1a76-74b2-4846-a9ee-7df99b00856a 254:0 0 476.4G 0 crypt / $ lsblk --fs NAME FSTYPE FSVER LABEL UUID FSAVAIL FSUSE% MOUNTPOINT loop0 squash 4.0 0 100% /var/lib/s loop1 squash 4.0 0 100% /var/lib/s loop2 squash 4.0 0 100% /var/lib/s loop3 squash 4.0 0 100% /var/lib/s loop4 squash 4.0 0 100% /var/lib/s loop5 squash 4.0 0 100% /var/lib/s loop6 squash 4.0 0 100% /var/lib/s loop7 squash 4.0 0 100% /var/lib/s loop8 squash 4.0 0 100% /var/lib/s loop9 squash 4.0 0 100% /var/lib/s nvme0n1 │ ├─nvme0n1p1 │ vfat FAT32 B12F-419C 551.2M 8% /boot/efi └─nvme0n1p2 crypto 1 d57f1a76-74b2-4846-a9ee-7df99b00856a └─luks-d57f1a76-74b2-4846-a9ee-7df99b00856a ext4 1.0 01657f93-ac1a-465b-8913-b526eb411b94 224.7G 47% / $ sudo dmsetup ls --target crypt luks-d57f1a76-74b2-4846-a9ee-7df99b00856a (254, 0) $ sudo lvmdiskscan /dev/nvme0n1 [ <476.94 GiB] /dev/loop0 [ 54.64 MiB] /dev/mapper/luks-d57f1a76-74b2-4846-a9ee-7df99b00856a [ 476.35 GiB] /dev/nvme0n1p1 [ 600.00 MiB] /dev/loop1 [ 160.15 MiB] /dev/nvme0n1p2 [ 476.35 GiB] /dev/loop2 [ <156.68 MiB] /dev/loop3 [ 91.31 MiB] /dev/loop4 [ 54.62 MiB] /dev/loop6 [ <91.37 MiB] /dev/loop7 [ <44.18 MiB] /dev/loop8 [ 44.88 MiB] 1 disk 11 partitions 0 LVM physical volume whole disks 0 LVM physical volumes Anyway, I tried running the benchmark with the following set in ~/.S-config.sh: PART=/dev/dm-0 BACKING_DEVS=nvme0n1 HIGH_LEV_DEV=dm-0 which I'm not sure is correct... and commenting out the following line: https://github.com/Algodev-github/S/blob/34f913259b1b9c1729cd08f05ff0277d7f1e2738/process_config.sh#L97 Results: results.tar.gz — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

from s.

teohhanhui commented on August 21, 2024

Here is the output if I reset everything:

$ run_multiple_benchmarks/test_responsiveness.sh
[sudo] password for teohhanhui: 
No user config found in /home/teohhanhui, copying default config
Not enough free space for test files in /home/teohhanhui/projects/Algodev-github/S/run_multiple_benchmarks/../workfiles: I need at least 500MB

Here's some additional information I've just figured out:

S/process_config.sh

Lines 263 to 265 in 34f9132

 if [[ "$PART" == "" ]]; then 

 PART=$(find_partition_for_dir $BASE_DIR) 

 fi

gives /dev/mapper/luks-d57f1a76-74b2-4846-a9ee-7df99b00856a, which is actually correct. It seems to be the ext4 partition on top of LUKS?

However,

S/process_config.sh

Lines 267 to 269 in 34f9132

 if [[ "$(df | egrep $PART)" == "" ]]; then # it must be /dev/root 

 PART=/dev/root 

 fi

tries to match from the df output and obviously fails.

from s.

Algodev-github commented on August 21, 2024

Il giorno 6 mar 2020, alle ore 18:30, Teoh Han Hui ***@***.***> ha scritto: Here is the output if I reset everything: https://gist.github.com/teohhanhui/4ff932382f58b389534f2ab847182074

Ok. So this is with default configuration. Does it work if you set TEST_DEV as I suggested, in ~/.S-config?

Here's some additional information I've just figured out: https://github.com/Algodev-github/S/blob/34f913259b1b9c1729cd08f05ff0277d7f1e2738/process_config.sh#L263-L265 gives /dev/mapper/luks-d57f1a76-74b2-4846-a9ee-7df99b00856a, which is actually correct. It seems to be the ext4 partition on top of LUKS?

Or below LUKS :) But yes, it seems correct.

However, https://github.com/Algodev-github/S/blob/34f913259b1b9c1729cd08f05ff0277d7f1e2738/process_config.sh#L267-L269 tries to match from the df output and obviously fails.

Yes, because this simple parsing code has not been written with LUKS in mind. So, even if it may have worked also for LUKS, with a little luck, it is no surprise that it fails. Thanks, Paolo

…

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

from s.

teohhanhui commented on August 21, 2024

Does it work if you set TEST_DEV as I suggested, in ~/.S-config?

Unfortunately not. This is what I get instead:

$ run_multiple_benchmarks/test_responsiveness.sh
[sudo] password for teohhanhui: 
Not enough free space for test files in /var/lib/S: I need at least 500MB

I added logging after this line:

S/process_config.sh

Line 347 in 34f9132

DISK=$(lsblk -o TYPE /dev/$TEST_DEV | egrep disk)

and this is what I get:

TEST_DEV=nvme0n1
TEST_PARTITION=nvme0n1p2
DISK=disk

🤷‍♂️

from s.

paolo-github commented on August 21, 2024

Il giorno 6 mar 2020, alle ore 18:44, Teoh Han Hui ***@***.***> ha scritto: Does it work if you set TEST_DEV as I suggested, in ~/.S-config? Unfortunately not. This is what I get instead: $ run_multiple_benchmarks/test_responsiveness.sh [sudo] password for teohhanhui: Not enough free space for test files in /var/lib/S: I need at least 500MB

Do you have at least 500MB of free space in the partition containing /var/lib/S? Is that directory in the right test partition?

I added logging after this line: https://github.com/Algodev-github/S/blob/34f913259b1b9c1729cd08f05ff0277d7f1e2738/process_config.sh#L347 and this is what I get: TEST_DEV=nvme0n1 DISK=disk

This is correct. But it doesn't tell us the partition. The partition suffix has certainly been stripped from the string you set, and the partition name has been put in a dedicated parameter (TEST_PARTITION IIRC). Thanks, Paolo

…

🤷‍♂ — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

from s.

teohhanhui commented on August 21, 2024

Do you have at least 500MB of free space in the partition containing /var/lib/S? Is that directory in the right test partition?

Yes and yes.

Looks like this line is still the culprit:

S/process_config.sh

Line 401 in 34f9132

if [[ "$(df | egrep $PART)" == "" ]]; then # it must be /dev/root

as it resets PART to /dev/root.

from s.

paolo-github commented on August 21, 2024

Ok once again, designed without LUKS :( If you want to invest a little of your time extending it, I'd be glad to help you!

…

Il giorno 6 mar 2020, alle ore 18:58, Teoh Han Hui ***@***.***> ha scritto: Do you have at least 500MB of free space in the partition containing /var/lib/S? Is that directory in the right test partition? Yes and yes. Looks like this line is still the culprit: https://github.com/Algodev-github/S/blob/34f913259b1b9c1729cd08f05ff0277d7f1e2738/process_config.sh#L401 as it resets PART to /dev/root. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

from s.

teohhanhui commented on August 21, 2024

Turns out it was a very simple fix. PR incoming!

from s.

teohhanhui commented on August 21, 2024

What I've noticed in the results is how the throughput gets reduced to ~30%, but then the latency is also reduced to ~30% (for the 10r-seq workload).

results.tar.gz

from s.

Algodev-github commented on August 21, 2024

In the line about latency with 10r-seq I read (unit of measure is seconds, for possible other readers of this thread): # Workload mq-deadline kyber bfq none 10r-seq 2.17575 2.18075 0.736 2.10975 This doesn't seem a 30% reduction, but a reduction of (2.10975-0.736)/2.10975 = .65 = 65% compared with the best performing alternative (none). It is unavoidable that this reduces throughput, because the random workload for starting up gnome-terminal yields a much lower throughput, and only BFQ succeeds in serving that workload so much (to minimize the start-up of gnome-terminal, as minimizing any latency is one of the main goals of BFQ). The other schedulers go on serving the ten sequential readers most of the time, which of course boosts throughput.

…

Il giorno 6 mar 2020, alle ore 19:37, Teoh Han Hui ***@***.***> ha scritto: What I've noticed in the results is how the throughput gets reduced to ~30%, but then the latency is also reduced to ~30% (for the 10r-seq workload). results.tar.gz — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

from s.

Support for NVMe devices / dm-crypt / LUKS / LVM about s HOT 10 CLOSED

Comments (10)

Related Issues (4)

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent

	if [[ "$PART" == "" ]]; then
	PART=$(find_partition_for_dir $BASE_DIR)
	fi

	if [[ "$(df \| egrep $PART)" == "" ]]; then # it must be /dev/root
	PART=/dev/root
	fi