Search Results

What you've done seems to me like it should work. Nothing jumps out at me as being incorrect. It could be a permissions issue; perhaps not.

What I would do is to get it working first by opening up...

Just guessing...

1. Possible typo in /etc/dfs/dfstab: share -F nfs -o rw /mount_point1
2. Possibly wrong configuration order: NFS daemon/server was started before the modification in dfstab, thus...

Well what is the mount point?
If it is a subdirectory of where you are when trying to mount, that could be the sort of expected message...

In UNIX you can NFS mount the directory from the server to the client.
Consult the man pages for setting up the NFS server and mounting to the client.

You don't tell us what format the backup is in, but if it's a file that you need to read from, then you can go with the suggestion from blackrageous above, subject to firewall rules and any...

Step values like /5 are not allowed in Unix crontab.
You need to give the full list 0,5,10,15,20,25,30,35,40,45,50,55
--
The list tells the exact times; it is obvious how to run them 1 minute...

You must be in the zone... Still showing the same zone means you havent changed environment, how are you login to the different zones? Using ssh?

VSZ shows claimed/reserved/overcommitted memory.
RSS shows resident/initialized memory.

Usually RSS and pmem are more relevant. Check yourself.
ps -u myuser -o pid,vsz,rss,pmem,args
Follow the...

It's not that simple. Oracle in particular uses lots and lots of shared memory. A good amount for each process is likely to be the same memory, shared between all of them -- just making a blind...

$ man ps

...
sz SZ size in physical pages of the core image of the
process. This includes text, data, and stack
...

Below a nagios plugin:
#!/bin/sh
#
# Warn if a user consumes too much memory (RSS)
# ARG 1: limit (MB)
#

set -f
export PATH
PATH=/bin:/usr/bin:/usr/sbin:/sbin

limit=${1:-5700}
check=`ps...

prstat -a

You may find it useful to run a very simple one-off cron which will show you the cron environment.

e.g.

set >/tmp/mylog.log

Then compare the output from the command "set" when running from...

If there was no crontab, crontab -l should silently do nothing, yes? But they get "can't open", which might be a permissions issue instead.

Actual crontabs live in

/usr/spool/cron/crontabs/<username>

so for user root there should be a file

/usr/spool/cron/crontabs/root

Not all users need to have a crontab file...

It cant open something inexistant...
You will have to write one from scratch...
call it mycronfile and when done you launch it by typing:

crontab mycronfile

If this were a Linux box the problem would be obvious, you need to add them to the cron group, but I'm uncertain if Solaris cron works the same way.

@Aia......yes, that is the normal mode of operation BUT (as Corona688 says), if the access rights anywhere on that path have been changed to deny the user access, then no go. It all depends whether...

Don't know what to make of this information in this guide (http://docs.oracle.com/cd/E23824_01/html/821-1451/sysrescron-23.html#scrolltoc), but it is the same error

Only if you place them in a script.

You can group them with { ... ; } or ( ... ) or you can use exec.

Regards,
Alister

Yes
exec >> /tmp/output.log 2>&1
Put this line at the beginning of the script and all the following commands will redirect their output to the mentioned file

This question made me worry when I saw it, and that's why -- take it to its natural conclusion and it stops working. Shell doesn't work that way... It will split spaces, but it won't start over...

if [ "$db" = "$prod" ]
then
CONNECT="sqlplus ser/password@dbname"
elif [ "$db" = "$test" ]
then
CONNECT="sqlplus ser/password@dbname"
else
exit
fi
$CONNECT

If the files are sorted, try:
comm -3 file1 file2