vnstat 2.6: Fix for lfs

Message ID 20200411143848.23912-1-matthias.fischer@ipfire.org
State Dropped
Headers
Series vnstat 2.6: Fix for lfs |

Commit Message

Matthias Fischer April 11, 2020, 2:38 p.m. UTC
  Removed 'sleep 2'

Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
---
 src/initscripts/system/vnstat | 2 --
 1 file changed, 2 deletions(-)
  

Comments

Michael Tremer April 11, 2020, 3:28 p.m. UTC | #1
Hello,

Thanks for the patch, but this only solves one of my concerns.

What about printing the colourful exit status more than once?

Best,
-Michael

> On 11 Apr 2020, at 15:38, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
> 
> Removed 'sleep 2'
> 
> Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
> ---
> src/initscripts/system/vnstat | 2 --
> 1 file changed, 2 deletions(-)
> 
> diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
> index bcc19c3ab..a21905d75 100755
> --- a/src/initscripts/system/vnstat
> +++ b/src/initscripts/system/vnstat
> @@ -20,14 +20,12 @@ case "$1" in
> 
> 		boot_mesg "Starting vnstatd..."
> 		loadproc /usr/sbin/vnstatd -d --alwaysadd
> -		sleep 2
> 		evaluate_retval
> 		;;
> 
> 	stop)
> 		boot_mesg "Stopping vnstatd..."
> 		killproc /usr/sbin/vnstatd
> -		sleep 2
> 		evaluate_retval
> 
> 		umount_ramdisk "${VNSTATLOG}"
> -- 
> 2.18.0
>
  
Matthias Fischer April 11, 2020, 4:39 p.m. UTC | #2
Hi.

On 11.04.2020 17:28, Michael Tremer wrote:
> Hello,
> 
> Thanks for the patch, but this only solves one of my concerns.

You mean the exit status for the 'stop'-section?

> What about printing the colourful exit status more than once?

Done... ;-)

> Best,
> -Michael
> 
>> On 11 Apr 2020, at 15:38, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>> 
>> Removed 'sleep 2'
>> 
>> Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
>> ---
>> src/initscripts/system/vnstat | 2 --
>> 1 file changed, 2 deletions(-)
>> 
>> diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
>> index bcc19c3ab..a21905d75 100755
>> --- a/src/initscripts/system/vnstat
>> +++ b/src/initscripts/system/vnstat
>> @@ -20,14 +20,12 @@ case "$1" in
>> 
>> 		boot_mesg "Starting vnstatd..."
>> 		loadproc /usr/sbin/vnstatd -d --alwaysadd
>> -		sleep 2
>> 		evaluate_retval
>> 		;;
>> 
>> 	stop)
>> 		boot_mesg "Stopping vnstatd..."
>> 		killproc /usr/sbin/vnstatd
>> -		sleep 2
>> 		evaluate_retval
>> 
>> 		umount_ramdisk "${VNSTATLOG}"
>> -- 
>> 2.18.0
>> 
>
  
Michael Tremer April 11, 2020, 6:03 p.m. UTC | #3
Hi,

> On 11 Apr 2020, at 17:39, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
> 
> Hi.
> 
> On 11.04.2020 17:28, Michael Tremer wrote:
>> Hello,
>> 
>> Thanks for the patch, but this only solves one of my concerns.
> 
> You mean the exit status for the 'stop'-section?

No, everywhere.

> 
>> What about printing the colourful exit status more than once?
> 
> Done... ;-)
> 
>> Best,
>> -Michael
>> 
>>> On 11 Apr 2020, at 15:38, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>> 
>>> Removed 'sleep 2'
>>> 
>>> Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
>>> ---
>>> src/initscripts/system/vnstat | 2 --
>>> 1 file changed, 2 deletions(-)
>>> 
>>> diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
>>> index bcc19c3ab..a21905d75 100755
>>> --- a/src/initscripts/system/vnstat
>>> +++ b/src/initscripts/system/vnstat
>>> @@ -20,14 +20,12 @@ case "$1" in
>>> 
>>> 		boot_mesg "Starting vnstatd..."
>>> 		loadproc /usr/sbin/vnstatd -d --alwaysadd
>>> -		sleep 2
>>> 		evaluate_retval
>>> 		;;
>>> 
>>> 	stop)
>>> 		boot_mesg "Stopping vnstatd..."
>>> 		killproc /usr/sbin/vnstatd
>>> -		sleep 2
>>> 		evaluate_retval
>>> 
>>> 		umount_ramdisk "${VNSTATLOG}"
>>> -- 
>>> 2.18.0
>>> 
>> 
>
  
Matthias Fischer April 11, 2020, 6:17 p.m. UTC | #4
Hi,

On 11.04.2020 20:03, Michael Tremer wrote:
> Hi,
> 
>> On 11 Apr 2020, at 17:39, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>> 
>> Hi.
>> 
>> On 11.04.2020 17:28, Michael Tremer wrote:
>>> Hello,
>>> 
>>> Thanks for the patch, but this only solves one of my concerns.
>> 
>> You mean the exit status for the 'stop'-section?
> 
> No, everywhere.

-v please.

'evaluate_retval' is now four times in, status output has been added.

Anything else?

> 
>> 
>>> What about printing the colourful exit status more than once?
>> 
>> Done... ;-)
>> 
>>> Best,
>>> -Michael
>>> 
>>>> On 11 Apr 2020, at 15:38, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>>> 
>>>> Removed 'sleep 2'
>>>> 
>>>> Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
>>>> ---
>>>> src/initscripts/system/vnstat | 2 --
>>>> 1 file changed, 2 deletions(-)
>>>> 
>>>> diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
>>>> index bcc19c3ab..a21905d75 100755
>>>> --- a/src/initscripts/system/vnstat
>>>> +++ b/src/initscripts/system/vnstat
>>>> @@ -20,14 +20,12 @@ case "$1" in
>>>> 
>>>> 		boot_mesg "Starting vnstatd..."
>>>> 		loadproc /usr/sbin/vnstatd -d --alwaysadd
>>>> -		sleep 2
>>>> 		evaluate_retval
>>>> 		;;
>>>> 
>>>> 	stop)
>>>> 		boot_mesg "Stopping vnstatd..."
>>>> 		killproc /usr/sbin/vnstatd
>>>> -		sleep 2
>>>> 		evaluate_retval
>>>> 
>>>> 		umount_ramdisk "${VNSTATLOG}"
>>>> -- 
>>>> 2.18.0
>>>> 
>>> 
>> 
>
  
Michael Tremer April 11, 2020, 6:21 p.m. UTC | #5
> On 11 Apr 2020, at 19:17, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
> 
> Hi,
> 
> On 11.04.2020 20:03, Michael Tremer wrote:
>> Hi,
>> 
>>> On 11 Apr 2020, at 17:39, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>> 
>>> Hi.
>>> 
>>> On 11.04.2020 17:28, Michael Tremer wrote:
>>>> Hello,
>>>> 
>>>> Thanks for the patch, but this only solves one of my concerns.
>>> 
>>> You mean the exit status for the 'stop'-section?
>> 
>> No, everywhere.
> 
> -v please.

LOL. Good one :)

So, you are launching vnstatd by calling loadproc.

That function will start the process and print “OK” or “ERROR” in a colour on the console.

Calling evaluate_retval (which should be called in loadproc, too) will then print OK again.

Did it not do that when you tested it?

-Michael

> 
> 'evaluate_retval' is now four times in, status output has been added.
> 
> Anything else?
> 
>> 
>>> 
>>>> What about printing the colourful exit status more than once?
>>> 
>>> Done... ;-)
>>> 
>>>> Best,
>>>> -Michael
>>>> 
>>>>> On 11 Apr 2020, at 15:38, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>>>> 
>>>>> Removed 'sleep 2'
>>>>> 
>>>>> Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
>>>>> ---
>>>>> src/initscripts/system/vnstat | 2 --
>>>>> 1 file changed, 2 deletions(-)
>>>>> 
>>>>> diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
>>>>> index bcc19c3ab..a21905d75 100755
>>>>> --- a/src/initscripts/system/vnstat
>>>>> +++ b/src/initscripts/system/vnstat
>>>>> @@ -20,14 +20,12 @@ case "$1" in
>>>>> 
>>>>> 		boot_mesg "Starting vnstatd..."
>>>>> 		loadproc /usr/sbin/vnstatd -d --alwaysadd
>>>>> -		sleep 2
>>>>> 		evaluate_retval
>>>>> 		;;
>>>>> 
>>>>> 	stop)
>>>>> 		boot_mesg "Stopping vnstatd..."
>>>>> 		killproc /usr/sbin/vnstatd
>>>>> -		sleep 2
>>>>> 		evaluate_retval
>>>>> 
>>>>> 		umount_ramdisk "${VNSTATLOG}"
>>>>> -- 
>>>>> 2.18.0
>>>>> 
>>>> 
>>> 
>> 
>
  
Matthias Fischer April 11, 2020, 6:22 p.m. UTC | #6
Hi,

On 11.04.2020 20:03, Michael Tremer wrote:
> Hi,
> 
>> On 11 Apr 2020, at 17:39, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>> 
>> Hi.
>> 
>> On 11.04.2020 17:28, Michael Tremer wrote:
>>> Hello,
>>> 
>>> Thanks for the patch, but this only solves one of my concerns.
>> 
>> You mean the exit status for the 'stop'-section?
> 
> No, everywhere.
> 
>> ...

Current version contains 'evaluate_retval' in all start/stop actions
plus output for 'status':

=>
https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/vnstat;h=c1bb2942a897e0a4acab344c8e2c128d4542f958;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff

Please check if this is ok now.

Best,
Matthias
  
Michael Tremer April 11, 2020, 6:26 p.m. UTC | #7
Hi,

Yes, but this script does not contain any evaluate_retval calls after loadproc/killproc:

https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/lvmetad;h=fdae39fd9c74e3950495ee54a1f6762af951fab4;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff

-Michael

> On 11 Apr 2020, at 19:22, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
> 
> Hi,
> 
> On 11.04.2020 20:03, Michael Tremer wrote:
>> Hi,
>> 
>>> On 11 Apr 2020, at 17:39, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>> 
>>> Hi.
>>> 
>>> On 11.04.2020 17:28, Michael Tremer wrote:
>>>> Hello,
>>>> 
>>>> Thanks for the patch, but this only solves one of my concerns.
>>> 
>>> You mean the exit status for the 'stop'-section?
>> 
>> No, everywhere.
>> 
>>> ...
> 
> Current version contains 'evaluate_retval' in all start/stop actions
> plus output for 'status':
> 
> =>
> https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/vnstat;h=c1bb2942a897e0a4acab344c8e2c128d4542f958;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff
> 
> Please check if this is ok now.
> 
> Best,
> Matthias
>
  
Matthias Fischer April 11, 2020, 6:40 p.m. UTC | #8
Hi,

On 11.04.2020 20:21, Michael Tremer wrote:
> 
> 
>> On 11 Apr 2020, at 19:17, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>> 
>> Hi,
>> 
>> On 11.04.2020 20:03, Michael Tremer wrote:
>>> Hi,
>>> 
>>>> On 11 Apr 2020, at 17:39, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>>> 
>>>> Hi.
>>>> 
>>>> On 11.04.2020 17:28, Michael Tremer wrote:
>>>>> Hello,
>>>>> 
>>>>> Thanks for the patch, but this only solves one of my concerns.
>>>> 
>>>> You mean the exit status for the 'stop'-section?
>>> 
>>> No, everywhere.
>> 
>> -v please.
> 
> LOL. Good one :)

:-)

> So, you are launching vnstatd by calling loadproc.

Yep.

> That function will start the process and print “OK” or “ERROR” in a colour on the console.

I noticed that, but...

> Calling evaluate_retval (which should be called in loadproc, too) will then print OK again.

I saw that, too, but didn't think of ALL actions in that script.

> Did it not do that when you tested it?

Nope. Doing it that way was a bit unfamiliar but I'll get used to it.

I had to take a look at the other inits and 'functions.pl'.

Now I finally know where the colorful "OKs" come from...you're never too
old... ;-))

Does that last version fit your needs?

Current link is:
https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/vnstat;h=c1bb2942a897e0a4acab344c8e2c128d4542f958;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff

Best,
Matthias

> -Michael
> 
>> 
>> 'evaluate_retval' is now four times in, status output has been added.
>> 
>> Anything else?
>> 
>>> 
>>>> 
>>>>> What about printing the colourful exit status more than once?
>>>> 
>>>> Done... ;-)
>>>> 
>>>>> Best,
>>>>> -Michael
>>>>> 
>>>>>> On 11 Apr 2020, at 15:38, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
>>>>>> 
>>>>>> Removed 'sleep 2'
>>>>>> 
>>>>>> Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
>>>>>> ---
>>>>>> src/initscripts/system/vnstat | 2 --
>>>>>> 1 file changed, 2 deletions(-)
>>>>>> 
>>>>>> diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
>>>>>> index bcc19c3ab..a21905d75 100755
>>>>>> --- a/src/initscripts/system/vnstat
>>>>>> +++ b/src/initscripts/system/vnstat
>>>>>> @@ -20,14 +20,12 @@ case "$1" in
>>>>>> 
>>>>>> 		boot_mesg "Starting vnstatd..."
>>>>>> 		loadproc /usr/sbin/vnstatd -d --alwaysadd
>>>>>> -		sleep 2
>>>>>> 		evaluate_retval
>>>>>> 		;;
>>>>>> 
>>>>>> 	stop)
>>>>>> 		boot_mesg "Stopping vnstatd..."
>>>>>> 		killproc /usr/sbin/vnstatd
>>>>>> -		sleep 2
>>>>>> 		evaluate_retval
>>>>>> 
>>>>>> 		umount_ramdisk "${VNSTATLOG}"
>>>>>> -- 
>>>>>> 2.18.0
>>>>>> 
>>>>> 
>>>> 
>>> 
>> 
>
  
Matthias Fischer April 11, 2020, 6:41 p.m. UTC | #9
On 11.04.2020 20:26, Michael Tremer wrote:
> Hi,
> 
> Yes, but this script does not contain any evaluate_retval calls after loadproc/killproc:
> 
> https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/lvmetad;h=fdae39fd9c74e3950495ee54a1f6762af951fab4;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff
> 
> ...

Current link is (I updated the one above!):
https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/vnstat;h=c1bb2942a897e0a4acab344c8e2c128d4542f958;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff

Best,
Matthias
  
Michael Tremer April 12, 2020, 2:15 p.m. UTC | #10
Hi,

> On 11 Apr 2020, at 19:41, Matthias Fischer <matthias.fischer@ipfire.org> wrote:
> 
> On 11.04.2020 20:26, Michael Tremer wrote:
>> Hi,
>> 
>> Yes, but this script does not contain any evaluate_retval calls after loadproc/killproc:
>> 
>> https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/lvmetad;h=fdae39fd9c74e3950495ee54a1f6762af951fab4;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff
>> 
>> ...
> 
> Current link is (I updated the one above!):
> https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/vnstat;h=c1bb2942a897e0a4acab344c8e2c128d4542f958;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff

So in this script, when you remove line 23 and 33, you will still see OK.

That is being printed in loadproc. evaluate_retval will now always overwrite what loadproc has printed to the right hand side of the console.

We want to see any errors here :)

-Michael

> 
> Best,
> Matthias
  
Matthias Fischer April 12, 2020, 5:22 p.m. UTC | #11
Hi,

On 12.04.2020 16:15, Michael Tremer wrote:
>> ...
>> Current link is (I updated the one above!):
>> https://git.ipfire.org/?p=people/mfischer/ipfire-2.x.git;a=blob;f=src/initscripts/system/vnstat;h=c1bb2942a897e0a4acab344c8e2c128d4542f958;hb=59b3afd07f55b93f4c96bfe63d0506494a46b5ff
>
> So in this script, when you remove line 23 and 33, you will still see OK.

To make this clear to me:

First:
Line 23 contains 'evaluate_retval' after 'loadproc...vnstat'. This is
counterproductive, because...:

> That is being printed in loadproc. evaluate_retval will now always overwrite what loadproc has printed to the right hand side of the console.

Ok, understood. I'll delete line 23.

Second:
'evaluate_retval' in line 33 (umount_ramdisk ...) is similar with line
23. Here the output of evaluate_retval' would overwrite error messages
regarding the unmounting of the ramdisk.(?)

But: in this case I can't see where the [OK] should come from (see
above: "...you will still see OK") when I remove line 33, because I'll
see only the output of 'boot_msg'!?

Would be ok for me, I'll delete line 33, too - I'm just wondering.

> We want to see any errors here

Me, too. ;-)

Best,
Matthias
  

Patch

diff --git a/src/initscripts/system/vnstat b/src/initscripts/system/vnstat
index bcc19c3ab..a21905d75 100755
--- a/src/initscripts/system/vnstat
+++ b/src/initscripts/system/vnstat
@@ -20,14 +20,12 @@  case "$1" in
 
 		boot_mesg "Starting vnstatd..."
 		loadproc /usr/sbin/vnstatd -d --alwaysadd
-		sleep 2
 		evaluate_retval
 		;;
 
 	stop)
 		boot_mesg "Stopping vnstatd..."
 		killproc /usr/sbin/vnstatd
-		sleep 2
 		evaluate_retval
 
 		umount_ramdisk "${VNSTATLOG}"