calise-testing

[Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Yes, i know you already knew it.
But only to be sure, it happens sometimes that calised stops doing its
work.
calised -d will say that sleep time between captures is a normal number,
but then, after those seconds, it won't start a capture.
calised -c works instead.
No errors in log...if i reboot everything is ok. It only happens sometimes,
and probably i should restart calised to have it work again.
What is it related to?
:) thanks!

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Il 09/07/2012 19:26, Developement versions test results and discussion 
ha scritto:
> Yes, i know you already knew it.
> But only to be sure, it happens sometimes that calised stops doing its 
> work.
> calised -d will say that sleep time between captures is a normal 
> number, but then, after those seconds, it won't start a capture.
> calised -c works instead.
> No errors in log...if i reboot everything is ok. It only happens 
> sometimes, and probably i should restart calised to have it work again.
> What is it related to?
> :) thanks!

Well, too few infos...
First of all make sure you have latest git master: GIT-e7176e3
Then, next time this happens, let me have the complete log (you can 
redirect the command below to file):

    cat `calised --dump-settings | grep logfile | aw'{print $NF}'`

and maybe also:

    calised --dump-all


Hope we're able to find why this happens.

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Il 09/07/2012 19:26, Developement versions test results and discussion 
ha scritto:
> Yes, i know you already knew it.
> But only to be sure, it happens sometimes that calised stops doing its 
> work.
> calised -d will say that sleep time between captures is a normal 
> number, but then, after those seconds, it won't start a capture.
> calised -c works instead.
> No errors in log...if i reboot everything is ok. It only happens 
> sometimes, and probably i should restart calised to have it work again.
> What is it related to?
> :) thanks!

More or less I've been able to reproduce your bug report. Until it's 
fixed you can do

    calised --restart

as workaround, keeps everything (service, pid, tempfiles and log) but 
restarts only the "capture" thread.

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Yes, i knew it :)
thanks!
ps: what's new in today update? (i know i ask you this everytime sorry!i'm
only curious )
and sorry if i did not answer you, but after i wrote the first mail, i did
a calised --restart and i haven't encountered that problem yet.
bye!

2012/7/11 Developement versions test results and discussion <
cal...@li...>

> Il 09/07/2012 19:26, Developement versions test results and discussion
> ha scritto:
> > Yes, i know you already knew it.
> > But only to be sure, it happens sometimes that calised stops doing its
> > work.
> > calised -d will say that sleep time between captures is a normal
> > number, but then, after those seconds, it won't start a capture.
> > calised -c works instead.
> > No errors in log...if i reboot everything is ok. It only happens
> > sometimes, and probably i should restart calised to have it work again.
> > What is it related to?
> > :) thanks!
>
> More or less I've been able to reproduce your bug report. Until it's
> fixed you can do
>
>     calised --restart
>
> as workaround, keeps everything (service, pid, tempfiles and log) but
> restarts only the "capture" thread.
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Calise-testing mailing list
> Cal...@li...
> https://lists.sourceforge.net/lists/listinfo/calise-testing
>

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Il 11/07/2012 19:46, Developement versions test results and discussion 
ha scritto:
> Yes, i knew it :)
> thanks!
> ps: what's new in today update? (i know i ask you this everytime 
> sorry!i'm only curious )
> and sorry if i did not answer you, but after i wrote the first mail, i 
> did a calised --restart and i haven't encountered that problem yet.
> bye!

Today I updated frequently because I'm going to release 0.4.0 very soon 
(probably tomorrow).
Version GIT-1d41599 should be (future) release 0.4.0. I fixed (or at 
least tried to) every critical or feature blocking bug.*

I'll ask (if you have time) to check if systemd script implementation is 
ok (it hasn't to be executable, right?).

Now, about changes, I'll love anyone asking me that, so don't be shy 
asking things or reporting bugs, if I have time I'll answer.
git shortlog dump (Wed Jul 11 2012):

    08:16:05 | few optimizations and capture anti-lock method implementation
    10:44:02 | addedd init and systemd scripts, changed calised command
    priority, fixed minor imprecision in setup.py and changed version
    (pre-release)
    17:16:21 | init_script pidof-related bugfix
    17:26:11 | QtGui export-data bugfix
    18:18:17 | final init_script bugfix, now works
    21:01:11 | fixed issue 3542639, workaround for issue 3541919

Poorly speaking I "fixed this and that", prepared everything for release 
and so on.

Btw, I'm (slowly, need more that 24h per day) updating the wiki (as 
0.4.0 is released, also the wiki will be). I don't remember the changing 
policies I put (probably registration is needed) but everyone should be 
able to change any page.
If you have suggestions please report to me, if you think something 
needs to be modified, modify it.
Of course if you want to complete/add sections you're absolutely welcome!
If you lost it and you don't want to search it on the forums: 
http://calise.sourceforge.net/mediawiki/


* I still have iss3541919 branch active on git but complete fix will 
came with 0.4.1 or farther. Right is threated as critical error.

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

I have no time now to write in your wiki, but i already had a look at it!
As soon as 0.4 is released, i'll read it again! very  nice btw!
Systemd .service mustn't be executable. May be, make the install script
outputs something like :
"Don't forget to run systemctl enable calised.service (you named it
calised,service or calise.service? i do not remember!) if you want calised
to be start as root with systemd!"
And in archlinux italian forum i wrote the .timer too, because systemd
makes use of aggressive parallelization of processes, and calised is a slow
process (because it needs access to webcam) and make it run during the boot
process (while lots of processes are started, and there is intense I/O
activity) make the whole boot time increase of 7-8'' or more. (at least for
me.)
Plus, i don't know when webcam can be accessed without any error (after the
right module has been loaded, for those who have an initramfs, i've not it,
so i can't tell you), and this thing can create others problem.
So, if you think i'm right, and if you followed me (hope so, sorry if i
couldn't explain me very well), please add the .timer file and everything
will be ok.
In that case, change the output i said before in:
"Don't forget to run systemctl enable calised.service or [reccommended]
calised.timer (25seconds after startup, more information man systemd.timer)
if you want calised to be start as root with systemd!"
Hope this can be helpful :)

2012/7/11 Developement versions test results and discussion <
cal...@li...>

> Il 11/07/2012 19:46, Developement versions test results and discussion
> ha scritto:
> > Yes, i knew it :)
> > thanks!
> > ps: what's new in today update? (i know i ask you this everytime
> > sorry!i'm only curious )
> > and sorry if i did not answer you, but after i wrote the first mail, i
> > did a calised --restart and i haven't encountered that problem yet.
> > bye!
>
> Today I updated frequently because I'm going to release 0.4.0 very soon
> (probably tomorrow).
> Version GIT-1d41599 should be (future) release 0.4.0. I fixed (or at
> least tried to) every critical or feature blocking bug.*
>
> I'll ask (if you have time) to check if systemd script implementation is
> ok (it hasn't to be executable, right?).
>
> Now, about changes, I'll love anyone asking me that, so don't be shy
> asking things or reporting bugs, if I have time I'll answer.
> git shortlog dump (Wed Jul 11 2012):
>
>     08:16:05 | few optimizations and capture anti-lock method
> implementation
>     10:44:02 | addedd init and systemd scripts, changed calised command
>     priority, fixed minor imprecision in setup.py and changed version
>     (pre-release)
>     17:16:21 | init_script pidof-related bugfix
>     17:26:11 | QtGui export-data bugfix
>     18:18:17 | final init_script bugfix, now works
>     21:01:11 | fixed issue 3542639, workaround for issue 3541919
>
> Poorly speaking I "fixed this and that", prepared everything for release
> and so on.
>
> Btw, I'm (slowly, need more that 24h per day) updating the wiki (as
> 0.4.0 is released, also the wiki will be). I don't remember the changing
> policies I put (probably registration is needed) but everyone should be
> able to change any page.
> If you have suggestions please report to me, if you think something
> needs to be modified, modify it.
> Of course if you want to complete/add sections you're absolutely welcome!
> If you lost it and you don't want to search it on the forums:
> http://calise.sourceforge.net/mediawiki/
>
>
> * I still have iss3541919 branch active on git but complete fix will
> came with 0.4.1 or farther. Right is threated as critical error.
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Calise-testing mailing list
> Cal...@li...
> https://lists.sourceforge.net/lists/listinfo/calise-testing
>

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Il 11/07/2012 22:30, Developement versions test results and discussion 
ha scritto:
> I have no time now to write in your wiki, but i already had a look at 
> it! As soon as 0.4 is released, i'll read it again! very  nice btw!
> Systemd .service mustn't be executable. May be, make the install 
> script outputs something like :
> "Don't forget to run systemctl enable calised.service (you named it 
> calised,service or calise.service? i do not remember!) if you want 
> calised to be start as root with systemd!"
> And in archlinux italian forum i wrote the .timer too, because systemd 
> makes use of aggressive parallelization of processes, and calised is a 
> slow process (because it needs access to webcam) and make it run 
> during the boot process (while lots of processes are started, and 
> there is intense I/O activity) make the whole boot time increase of 
> 7-8'' or more. (at least for me.)
> Plus, i don't know when webcam can be accessed without any error 
> (after the right module has been loaded, for those who have an 
> initramfs, i've not it, so i can't tell you), and this thing can 
> create others problem.
> So, if you think i'm right, and if you followed me (hope so, sorry if 
> i couldn't explain me very well), please add the .timer file and 
> everything will be ok.
> In that case, change the output i said before in:
> "Don't forget to run systemctl enable calised.service or 
> [reccommended] calised.timer (25seconds after startup, more 
> information man systemd.timer) if you want calised to be start as root 
> with systemd!"
> Hope this can be helpful :)

One of the bugs I fixed today should have been the one that caused the 
process to not start well if called before ConsoleKit loads. So, at 
least on my machine (with initscripts), calised can run just after dbus 
is loaded (if you can try with systemd, that will be *really* helpful, 
of course, after you tested put your "timer" back in).
Now, the timer it's a good thing but I won't include it in the package 
because the amount of time it's totally arbitrary, better should be 
writing a small page on the wiki about "running as system-wide service" 
at startup with either systemd or initd with all the hints and 
suggestions to have the perfect setup. Obviously including the "timer" 
thing and so on.

My idea for post-install text was that:

    initd and systemd startup scripts addedd

then everyone is free to search archlinux wikipages about daemons to 
know how to add them. Yes, I know, I *hate* post-install text (don't 
know why but that's it).

About boot time, for sure, calised, during the "startup", performs quite 
a lot of heavy operations. I mean it's more or less the same of loading 
cheese (actually way less) at system startup.

About the wiki, I'll split it up to several pages very soon, a big page 
with everything is *bad* (it's like writing a single function program).

Don't worry if you haven't enough time to test systemd script, the 
systemd/initd issue it's only archlinux related, so I can release 0.4.0 
anyway. I mean, I cannot make a startup script for every distribution, 
that's the packager duty (and as archlinux calise packager I'm getting 
it done).

Thank you very much for your cooperation.

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

I've now looked at new wiki page, congratulations, it is very well
explained and clear. Btw, i noticed startup script -> systemd is left
empty...as soon as i got time (may be today at about 19) i'll update it,
with .timer script and other hints (if i come up with them :) ).
I've had no time yet to upgrade to latest git and to test if calised can be
ran immediately after dbus (consider that on systemd you can't run
something immediately after something else, because of parallelization of
processes. You only can specify that a service must be started after
another, but this doesn't mean it will be started immediately after,  hope
i've been clear enough :) -> at least, as far as i know...may be i'm
missing something )
Btw i'll try today afternoon, yet about 19 o'clock.
I think you had a good idea about post-install text, that's clear enough.
Look at wiki bro! :)
Bye!

2012/7/11 Developement versions test results and discussion <
cal...@li...>

> Il 11/07/2012 22:30, Developement versions test results and discussion
> ha scritto:
> > I have no time now to write in your wiki, but i already had a look at
> > it! As soon as 0.4 is released, i'll read it again! very  nice btw!
> > Systemd .service mustn't be executable. May be, make the install
> > script outputs something like :
> > "Don't forget to run systemctl enable calised.service (you named it
> > calised,service or calise.service? i do not remember!) if you want
> > calised to be start as root with systemd!"
> > And in archlinux italian forum i wrote the .timer too, because systemd
> > makes use of aggressive parallelization of processes, and calised is a
> > slow process (because it needs access to webcam) and make it run
> > during the boot process (while lots of processes are started, and
> > there is intense I/O activity) make the whole boot time increase of
> > 7-8'' or more. (at least for me.)
> > Plus, i don't know when webcam can be accessed without any error
> > (after the right module has been loaded, for those who have an
> > initramfs, i've not it, so i can't tell you), and this thing can
> > create others problem.
> > So, if you think i'm right, and if you followed me (hope so, sorry if
> > i couldn't explain me very well), please add the .timer file and
> > everything will be ok.
> > In that case, change the output i said before in:
> > "Don't forget to run systemctl enable calised.service or
> > [reccommended] calised.timer (25seconds after startup, more
> > information man systemd.timer) if you want calised to be start as root
> > with systemd!"
> > Hope this can be helpful :)
>
> One of the bugs I fixed today should have been the one that caused the
> process to not start well if called before ConsoleKit loads. So, at
> least on my machine (with initscripts), calised can run just after dbus
> is loaded (if you can try with systemd, that will be *really* helpful,
> of course, after you tested put your "timer" back in).
> Now, the timer it's a good thing but I won't include it in the package
> because the amount of time it's totally arbitrary, better should be
> writing a small page on the wiki about "running as system-wide service"
> at startup with either systemd or initd with all the hints and
> suggestions to have the perfect setup. Obviously including the "timer"
> thing and so on.
>
> My idea for post-install text was that:
>
>     initd and systemd startup scripts addedd
>
> then everyone is free to search archlinux wikipages about daemons to
> know how to add them. Yes, I know, I *hate* post-install text (don't
> know why but that's it).
>
> About boot time, for sure, calised, during the "startup", performs quite
> a lot of heavy operations. I mean it's more or less the same of loading
> cheese (actually way less) at system startup.
>
> About the wiki, I'll split it up to several pages very soon, a big page
> with everything is *bad* (it's like writing a single function program).
>
> Don't worry if you haven't enough time to test systemd script, the
> systemd/initd issue it's only archlinux related, so I can release 0.4.0
> anyway. I mean, I cannot make a startup script for every distribution,
> that's the packager duty (and as archlinux calise packager I'm getting
> it done).
>
> Thank you very much for your cooperation.
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Calise-testing mailing list
> Cal...@li...
> https://lists.sourceforge.net/lists/listinfo/calise-testing
>

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Il 12/07/2012 13:36, Developement versions test results and discussion 
ha scritto:
> I've now looked at new wiki page, congratulations, it is very well 
> explained and clear. Btw, i noticed startup script -> systemd is left 
> empty...as soon as i got time (may be today at about 19) i'll update 
> it, with .timer script and other hints (if i come up with them :) ).
> I've had no time yet to upgrade to latest git and to test if calised 
> can be ran immediately after dbus (consider that on systemd you can't 
> run something immediately after something else, because of 
> parallelization of processes. You only can specify that a service must 
> be started after another, but this doesn't mean it will be started 
> immediately after,  hope i've been clear enough :) -> at least, as far 
> as i know...may be i'm missing something )
> Btw i'll try today afternoon, yet about 19 o'clock.
> I think you had a good idea about post-install text, that's clear 
> enough. Look at wiki bro! :)
> Bye!

Yup, you've been clear, btw I already know how systemd works (I tried it 
some time ago but I was too much used with initd and I actually don't 
care about bootup time) that said, is more or less what "@" do in initd, 
only that it's based on that (and it's way better than initd...).
 From what I've seen on the .service it seems to me that that 
"type=dbus" (or whatever it was) means "start after dbus".

Now, don't worry about systemd wiki or else, when you'd have time you'll 
do it (I would have done myself but I cannot test if what I write is 
correct); as said before that's "optional".

And again, every idea about the program and/or the wiki it's always welcome.

Thanks for your interest
Bye

Re: [Calise-testing] calised sometimes stops doing captures

[Developement v. t. r. a. d. <cal...@li...>]

Ok, i tested it now. It works with the .service unit file you include in
your package (without .timer), but it adds something like 12seconds to boot
time :)
I'm now registering at mediawiki and may be i'll include the .timer soon,
and any hints if i come up with them.
Thanks for your wonderful software ;)

2012/7/12 Developement versions test results and discussion <
cal...@li...>

> Il 12/07/2012 13:36, Developement versions test results and discussion
> ha scritto:
> > I've now looked at new wiki page, congratulations, it is very well
> > explained and clear. Btw, i noticed startup script -> systemd is left
> > empty...as soon as i got time (may be today at about 19) i'll update
> > it, with .timer script and other hints (if i come up with them :) ).
> > I've had no time yet to upgrade to latest git and to test if calised
> > can be ran immediately after dbus (consider that on systemd you can't
> > run something immediately after something else, because of
> > parallelization of processes. You only can specify that a service must
> > be started after another, but this doesn't mean it will be started
> > immediately after,  hope i've been clear enough :) -> at least, as far
> > as i know...may be i'm missing something )
> > Btw i'll try today afternoon, yet about 19 o'clock.
> > I think you had a good idea about post-install text, that's clear
> > enough. Look at wiki bro! :)
> > Bye!
>
> Yup, you've been clear, btw I already know how systemd works (I tried it
> some time ago but I was too much used with initd and I actually don't
> care about bootup time) that said, is more or less what "@" do in initd,
> only that it's based on that (and it's way better than initd...).
>  From what I've seen on the .service it seems to me that that
> "type=dbus" (or whatever it was) means "start after dbus".
>
> Now, don't worry about systemd wiki or else, when you'd have time you'll
> do it (I would have done myself but I cannot test if what I write is
> correct); as said before that's "optional".
>
> And again, every idea about the program and/or the wiki it's always
> welcome.
>
> Thanks for your interest
> Bye
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Calise-testing mailing list
> Cal...@li...
> https://lists.sourceforge.net/lists/listinfo/calise-testing
>