does remastering improve load time?

[P5music]

Hello,
I ask whether a remastered tinycore system with a certain set of extensions included  loads faster, even few seconds.
Thanks in advance

[tinypoodle]

Time required to read from medium is analogous to size.

[P5music]

Are you saying that remastering shrinks size or that time is more or less the same?

[maro]

I'd say any additions to the initrd will slow things down (which might not be noticeable if they are just a few small files). Likewise any extensions integrated into the ISO image will cost time for their loop-mounted installation (if they are onboot). As the term remastering is used for both things you might want to read another post where I tried to spell out the difference.

So, NO I don't think remastering will give you any faster boot times. TC is anyway the fastes booting Linux 2.6 system I know of. Well, unless you thought of actually reducing the inird. But then I' love to know what you plan to strip off. If you don't need an X server you should use MicroCore, which then does not need to extract the additional files. But in my perception this is not a huge amount of time a saved.

BTW, I remember someone once did an analysis of the TC boot times, unfortunately I can't find it right now, but IIRC there was not much "fat" that could be cut out.

[tinypoodle]

All file system support for FAT* and ext* is one thing which could be stripped if not required.
But I doubt difference timewise would be worth the effort, difference in RAM usage could be a better motivation.

[P5music]

At least I think a remastering would solve the waitusb problem during boot from sd-cards or pendrives.

[tinypoodle]

At least I think a remastering would solve the waitusb problem during boot from sd-cards or pendrives.

1. What exactly do you refer to by "waitusb problem" ?
2. In which way could remastering change what exactly?

[P5music]

waitusb=10 , 8 or 5 is too long for me so I think remastering can allow "automagical" load of extensions.

[tinypoodle]

Err... too long in relation to what?

This thread has some references to boot times and resource usage with different methods of remastering per se:
http://forum.tinycorelinux.net/index.php?topic=7002.0

I'd look at comparison of times for reading from different media as a subject on its own, but you hadn't mentioned such at the start of thread...

[curaga]

In the last few releases you can wait for a specific LABEL, instead of only some amount of seconds.

IIRC the last bootchart I made was on 2.9.

[P5music]

like waitusb=sda2 ?

[curaga]

http://forum.tinycorelinux.net/index.php?topic=5909.0

waitusb=30:LABEL=tinycore

wait max 30 seconds, stop earlier if label tinycore shows up. Also works for UUID.

[P5music]

Sorry, but I do not understand. I think I do not know what is LABEL and what "shows up" mean, in relation to waitusb delay.

[tinypoodle]

Code: [Select]

blkid /dev/*d*

[tinypoodle]

That's excellent actually, as with waitusb=30:LABEL= USB1 could be handled reliably while not wasting any time in case of USB2, all with same setting.  

Wondering though if there would be a way to use that in case where more than one USB device is required at boot time...