On 12/2/20 2:22 AM, Michael D. Setzer II wrote:
Question: Doesn't the dnf process use the cache feature of the disk??
Not really, since it's mostly writing new data. It's not going to avoid
the cache, but the use of the cache is completely up to the drive.
Know that the higher number is actually the speed of accessing the ram
buffer of the disk. But the lower number is the physical speed of the
disk.
I was slight wrong about that benchmark info, but it's still not a
useful measurement.
> hdparm -Tt /dev/sda
>
> /dev/sda:
> Timing cached reads: 10030 MB in 1.99 seconds = 5031.74 MB/sec
This has nothing to do with the drive at all. It's measuring the
throughput of the Linux buffering system. It doesn't touch the drive at
all.
> Timing buffered disk reads: 412 MB in 3.01 seconds = 136.93 MB/sec
This is a sequential read measurement and it's not a bad number, but
it's completely irrelevant to dnf. You need to measure write speed and
random read/write access times.
Just know that I started process at about after 11pm, and the
download only took about 25 minutes for the 5.4G of files it reported.
It was just the update.
Yes, it's the updating part that takes a long time because that's when
it's doing lots of random reads and writes of the disk.
Perhaps an SSD drive would change things.
It would drastically change things. They have virtually no seek time,
so it's all read/write speed and for reference, this is my laptop SSD:
Timing buffered disk reads: 1518 MB in 3.00 seconds = 505.41 MB/sec
A write speed test gave me 467 MB/s
_______________________________________________
users mailing list -- users@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to users-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/users@xxxxxxxxxxxxxxxxxxxxxxx
