[hobbit] FILE check says "No files checked", but things appear configured correctly

[Haertig, David F (Dave)]

I had errors reported against hobbitd sayinf "Oversize status message"
and "Oversize data/client message".

So I added:

MAXMSG_STATUS=1024
MAXMSG_CLIENT=1024
MAXMSG_DATA=1024

to hobbitserver.cfg right near the top of the file, after the initial
comment line (but I don't think placement in the file matters, you could
probably put the new lines at the bottom too).. 

-----Original Message-----
From: Lund, Holly [mailto:holly.lund at hq.doe.gov] 
Sent: Thursday, October 25, 2007 6:51 AM
To: hobbit at hswn.dk
Subject: RE: [hobbit] FILE check says "No files checked", but things
appear configured correctly


 Where did you put this in hobbitserver.cfg and what variable did you
use

I have this problem on multiple Solaris 8 servers



Holly Lund
301-903-1174
202-586-4431

-----Original Message-----
From: Haertig, David F (Dave) [mailto:haertig at avaya.com]
Sent: Wednesday, October 24, 2007 8:43 PM
To: hobbit at hswn.dk
Subject: RE: [hobbit] FILE check says "No files checked", but things
appear configured correctly

Found it, I think.  "Oversize message" in hobbitd test.  So it appears
that the entire message is available under the "client data" link, but
the MAXMSG* size settings determine how much of that client data Hobbit
will process?  When I said before that I could see my data in the
"client data" display, that data was at the end of the message.  I guess
it was getting truncated during processing.  When I added entries for
larger message sizes to hobbitserver.cfg my FILE test problem looks like
it went away.

________________________________

From: Haertig, David F (Dave) [mailto:haertig at avaya.com]
Sent: Wednesday, October 24, 2007 4:44 PM
To: hobbit at hswn.dk
Subject: [hobbit] FILE check says "No files checked", but things appear
configured correctly


I have a FILE test that is not working ... the webpage displays "No
files checked" on one of two identicallly monitored servers.  The second
server works just fine.
 
The two servers are configured like this in client-local.cfg on the
Hobbit host:
 
[spiexp1]
file:/tmp/dfh.test
log:/var/adm/messages:10240
[wcscbe5]
file:/tmp/dfh.test
log:/var/adm/messages:10240

And like this in hobbit-clients.cfg:
 
HOST=spiexp1,wcscbe5
        FILE    /tmp/dfh.test size>1K
        LOAD    40.0 70.0
        DISK    /expsys 95 98
        DISK    /expert 95 98
        DISK    /inadsb 90 95
        DISK    * IGNORE

I can see the meta-data for my test file, /tmp/dfh.test, in the "client
data" display for each server, so both of them apparently got the
message and are uploading the required data:
 
[file:/tmp/dfh.test]
type:100000 (file)
mode:644 (-rw-r--r--)
linkcount:1
owner:12319 (haertig)
group:906 (tsc)
size:0
clock:1193264890 (2007/10/24-16:28:10)
atime:1193257900 (2007/10/24-14:31:40)
ctime:1193257900 (2007/10/24-14:31:40)
mtime:1193257900 (2007/10/24-14:31:40)
 
But spiexp1 is saying "No files checked" whereas wcscbe5 is working just
as I'd expect.  Both servers were given two hours to settle out all the
upload/download handshaking stuff before I checked the results.  Is
there something I'm missing in my setup?  Both servers are configured
very similarly, but they're not identical twins.  They run Solaris 7
(yeah, don't tell me - I know it's old!).  Possibly related, wcscbe5
shows the output of the 'top' command on it's 'cpu' column on the
webpage, but spiexp1 does not (the graph is there for spiexp1, but not
the 'top' data, however spiexp1's 'top' data IS in the "client data" ...
just not on the displayed webpage).
 
Thanks !


To unsubscribe from the hobbit list, send an e-mail to
hobbit-unsubscribe at hswn.dk