Re: {bb} big brother monitoring a df that it shouldn't be

Scott Mohnkern <[email protected]> Tue, 11 Mar 2008 17:03:26 -0400
Newsgroups gmane.network.bb4.general
Message-ID <[email protected]>
This is a multi-part message in MIME format.

--Boundary_(ID_hyhE0tb3KySZcQ+RgkUM3g)
Content-type: text/plain; charset=ISO-8859-1; format=flowed
Content-transfer-encoding: 7BIT

Well, I think I figured it out.

Somewhere big brother is doing a:

sort +4 <filename>

and I need it to do a:

sort -k 4 <filename>

$BBHOME/etc/bbsys.sh has a

DFSORT="4"

But changing this to "-k 4"  generates:

/bin/sort: open failed: +-k 4: No such file or directory

So the question is, what do I need to change so instead of doing a sort 
+4 <filename> it does a sort -k 4 <filename>

Scott

Scott Mohnkern wrote:
> Actually, I may have found the problem, I took a look at $BBHOME/BBOUT 
> and there are lots of these:
>
> /bin/sort: open failed: +4: No such file or directory
>
>
> Of course that's not helpful, as I'm not sure what file it's not able 
> to find.
>
> Scott
>
>
> Philip Clark wrote:
>> On Mon, 2008-03-10 at 13:27 -0400, Scott Mohnkern wrote:
>>   
>>> I'm having a problem with bb a 1.6e1 client reporting on a partition 
>>> that it shouldn't be.
>>>
>>> It's acting like the tmp directory for the client (in this case 
>>> /usr/local/bb16e1/tmp) is a partition, and reporting it full (and its 
>>> not a partition at all, just a directory).
>>>
>>> adding the following to bb-dftab doesn't seem to help:
>>>
>>> localhost:/usr/local/bb16e1/tmp:101:101
>>>
>>> For the life of me, I can't figure out where it's deciding that 
>>> /usr/local/bb16e1/tmp is a partition, and that it needs to be monitored.
>>>
>>> Any suggestions?
>>>     
>>
>> Erm, 1.6e1? This would be archaeology then. I'll just fetch my trowel :)
>>
>> These "tmp filesystem full" messages are not generally the result of
>> examining the disks. A more complete translation would be, "I tried to
>> save some data to my tmp directory and all I got was an empty file."
>>
>> Two well know candidates are the Solaris "tail" bug and the altered
>> "sort" options:
>>
>>   http://support.bb4.com/archive/200405/msg00222.html
>>
>>   http://support.bb4.com/archive/200603/msg00170.html
>>
>> If neither of those looks likely, check through the client scripts
>> until you find the part that is generating the error message. If it
>> appears more than once, append an identifier to the message to isolate
>> it. Then you can try running the commands that precede it manually
>> to get a better idea of why they are failing.
>>
>> If you still can't locate the problem, let the list know. Unfortunately,
>> you would need to actually send me the client tarball (off list) as I
>> don't have anything nearly as old as that.
>>
>> Cheers, Phil.
>>
>>
>>
>> --
>> Real Programmers don't play tennis, or any other sport that
>> requires you to change clothes. Mountain climbing is OK, and real
>> programmers wear their climbing boots to work in case a mountain
>> should suddenly spring up in the middle of the machine room.
>>
>>
>> --
>> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=
>> To unsubscribe from this list, or to subscribe to the bb-digest list
>> send e-mail to mailto:[email protected] with unsubscribe bb -and/or-
>> subscribe bb-digest in the BODY of the message.
>>   

--Boundary_(ID_hyhE0tb3KySZcQ+RgkUM3g)
Content-type: text/html; charset=ISO-8859-1
Content-transfer-encoding: 7BIT

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Well, I think I figured it out.<br>
<br>
Somewhere big brother is doing a:<br>
<br>
sort +4 &lt;filename&gt;<br>
<br>
and I need it to do a:<br>
<br>
sort -k 4 &lt;filename&gt;<br>
<br>
$BBHOME/etc/bbsys.sh has a <br>
<br>
DFSORT="4"<br>
<br>
But changing this to "-k 4"&nbsp; generates:<br>
<br>
/bin/sort: open failed: +-k 4: No such file or directory<br>
<br>
So the question is, what do I need to change so instead of doing a sort
+4 &lt;filename&gt; it does a sort -k 4 &lt;filename&gt;<br>
<br>
Scott<br>
<br>
Scott Mohnkern wrote:
<blockquote cite="mid:[email protected]" type="cite">
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
Actually, I may have found the problem, I took a look at $BBHOME/BBOUT
and there are lots of these:<br>
  <br>
/bin/sort: open failed: +4: No such file or directory<br>
  <br>
  <br>
Of course that's not helpful, as I'm not sure what file it's not able
to find.<br>
  <br>
Scott<br>
  <br>
  <br>
Philip Clark wrote:
  <blockquote cite="mid:1205189342.6590.19.camel@drebin" type="cite">
    <pre wrap="">On Mon, 2008-03-10 at 13:27 -0400, Scott Mohnkern wrote:
  </pre>
    <blockquote type="cite">
      <pre wrap="">I'm having a problem with bb a 1.6e1 client reporting on a partition 
that it shouldn't be.

It's acting like the tmp directory for the client (in this case 
/usr/local/bb16e1/tmp) is a partition, and reporting it full (and its 
not a partition at all, just a directory).

adding the following to bb-dftab doesn't seem to help:

localhost:/usr/local/bb16e1/tmp:101:101

For the life of me, I can't figure out where it's deciding that 
/usr/local/bb16e1/tmp is a partition, and that it needs to be monitored.

Any suggestions?
    </pre>
    </blockquote>
    <pre wrap=""><!---->
Erm, 1.6e1? This would be archaeology then. I'll just fetch my trowel :)

These "tmp filesystem full" messages are not generally the result of
examining the disks. A more complete translation would be, "I tried to
save some data to my tmp directory and all I got was an empty file."

Two well know candidates are the Solaris "tail" bug and the altered
"sort" options:

  <a moz-do-not-send="true" class="moz-txt-link-freetext"
 href="http://support.bb4.com/archive/200405/msg00222.html">http://support.bb4.com/archive/200405/msg00222.html</a>

  <a moz-do-not-send="true" class="moz-txt-link-freetext"
 href="http://support.bb4.com/archive/200603/msg00170.html">http://support.bb4.com/archive/200603/msg00170.html</a>

If neither of those looks likely, check through the client scripts
until you find the part that is generating the error message. If it
appears more than once, append an identifier to the message to isolate
it. Then you can try running the commands that precede it manually
to get a better idea of why they are failing.

If you still can't locate the problem, let the list know. Unfortunately,
you would need to actually send me the client tarball (off list) as I
don't have anything nearly as old as that.

Cheers, Phil.



--
Real Programmers don't play tennis, or any other sport that
requires you to change clothes. Mountain climbing is OK, and real
programmers wear their climbing boots to work in case a mountain
should suddenly spring up in the middle of the machine room.


--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=
To unsubscribe from this list, or to subscribe to the bb-digest list
send e-mail to <a moz-do-not-send="true" class="moz-txt-link-freetext"
 href="mailto:[email protected]">mailto:[email protected]</a> with unsubscribe bb -and/or-
subscribe bb-digest in the BODY of the message.
  </pre>
  </blockquote>
</blockquote>
</body>
</html>

--Boundary_(ID_hyhE0tb3KySZcQ+RgkUM3g)--
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=
To unsubscribe from this list, or to subscribe to the bb-digest list
send e-mail to mailto:[email protected] with unsubscribe bb -and/or-
subscribe bb-digest in the BODY of the message.