(old)Puppy Linux Discussion Forum

This report a very interesting need for understanding that occurs with SLACKO usage.

Unexplained phenomenon

When the sytem is initially booted and tailored (before save-session shutdown) the system was using 92MB after SAMBA setup and use.
NOW, AFTER save-session to the LIve media and rebooting, all subsequent boots are ONLY using 65MB.

This is cause for great puzzlement.

Thus, I am asking the community to look over the 2 reports of SLACKO (one without any changes and the other with SAMBA and reboot).

Please, if you can, explain why the system behavior is using LESS RESOURCES after initial save-session to the CD than was used before the save session.

The system's speed is great. And its resource sharing with LAN users is fast, as well with SAMBA installed and running. BUT, the observation of RAM usage is not obvious, to me, why this has improved by simply adding the SAMBA subsystem.

Thanks in advance for any explanation anyone can provide.

Note: There are 2 text system reports in the attached compressed file.
1 report is SLACKO with SAMBA after reboot with save-session showing 64MB RAM use
2nd report is SLACKO where nothing was done except to run Hardinfo showing SLACKO pristine after desktop startup and FirstRUN.

gcmartin wrote:

When the system is initially booted and tailored (before save-session shutdown) the system was using 92MB after SAMBA setup and use. 
NOW, AFTER save-session to the LIve media and rebooting, all subsequent boots are ONLY using 65MB. 

A couple of thoughts:

1. Patent it quickly before Apple/MS kick down your door!

2. After setup & save Samba is routinely clearing a shared cache that was previously only cleared on shutdown.

3. The Samba daemons are now preventing something else from loading - error logs may show what.

What you may be seeing is how Puppy works.
With no save file, everything is done with memory and nothing is cleared, from memory, as programs work. A ram disk is used.
After save file is made. Memory can be flushed, periodically, and everything does not have to stay in memory.

Pupmode 12
What has happened here is that at bootup Puppy found the pup_save.3fs file is on a fast hard drive partition, so decided to mount it directly on the top layer. Thus, there is no tmpfs ramdisk intermediary. File pup_save.3fs is directly read and written to.

This scenario is very good for RAM-challenged PCs. The initial ramdisk, that is, the file initrd.gz that gets loaded first when Puppy boots, is still there in ram, mounted at /initrd, however it only uses about 1.9M.

After save file is made. Memory can be flushed, periodically, and everything does not have to stay in memory.

I think you might be confusing the issue.
Once you are running from a save file any files that are created in your Puppy system will be saved to disk, rather than kept in RAM. This includes any files that your browser caches, for example. But they are not kept in memory and then "flushed" to disk periodically - they are saved to the disk in the first place.
Where flushing is involved is if you boot from I think either a USB flash drive, or a multisession CD/DVD. In this case files are kept in RAM so that you can take the disk out if necessary, and to minimise writes.

disciple wrote:Once you are running from a save file any files that are created in your Puppy system will be saved to disk, rather than kept in RAM.

I'm no Puppy expert, so I'm probably wrong, but...
I had formed the idea that...

Puppy doesn't "save" back to the pupsave during a session...
It actually COPIES changes back to the pupsave...
And they are retained in RAM? NO?

gcmartin wrote:... AFTER save-session to the LIve media ...

What was the media? Is this a multisession Puppy CD or DVD, created after installing and configuring Samba?

disciple,

I agree, depends on what media is used, how save file use is handled.
I was showing Pupmode 12 and 13 together as example.
Guess it was a little confusing.
the main point is, the use of a save file is why the memory usage changed.

Live media operations means I am running a mult-session CDRW for saving the work I do on the PC. The save-session file is one that is written to the Live media (the CDRW) when I shutdown/reboot the PC.

I find this a very fascinating phenomenon. Thus, I am seeking some reasonable explanation on how this is occurring. I am really puzzled.

Last night, I had a similar thing surface with Exprimo. It went from 100+MB RAM at startup. I loaded SAMBA and JRE. And upon reboot, I fould that a startup RAM use is down to 85MB???

The Puppy systems in each case are "screaming fast", but, I have a phenomenon that I am wondering how this is occurring. Its not exactly what I expected by the additions fo 2 additional subsystem added to the running OS.

Any help or if I can provide info to better clarify, please help. Thanks in advance.