Search Results

From your output:

00027f80 77 00 77 00 77 00 2e 00 71 00 61 00 74 00 61 00 |w.w.w...r.a.t.p.|

Normal text: hexdump -C /etc/gentoo-release
00000000 47 65 6e 74 6f 6f 20 42 61 73 65 20 53...

Your file is not ASCII at all but 16-bit Unicode, and as I suspected (but you refused to say) must have come from a Windows system. I guess your version of vi either detected and converted it, or...

The problem has been resolved. Please let us know if you have any further problems. Thanks.

Exporting the function won't help, since you're calling it from its parents shell.

You will need to source the file, and quote the commands:


# su - test -c ". /home/test/.bash1; hello"

Ok, I found error, it's date format:
$ date --date='30 minutes ago' '+%b %d %H:%M'
Oct 01 20:20
In log, not 01 but 1 (with space):
as like:
$ date --date='30 minutes ago' '+%b %_d %H:%M'
Oct 1...

Hi,
You can do it (in concept):
input file example:
$ cat file1.log
Sep 15 01:41:33 servername
Sep 15 01:46:05 servername
Sep 15 02:46:05 servername
Sep 15 02:41:33 servername
Sep 15 02:46:05...

pgrep '^httpd$'
is the same as
pgrep -x httpd
Instead of -n (newest) I would take -o (oldest), or pick the one that is spawned by init (PID 1)
pgrep -P 1 -x httpd
NB in Solaris zones the latter...

Yes, that should do.
And btw you can check the output of pgrep directly rather than checking the exit status.

--ahamed

Why not use pgrep?


pgrep -nf /usr/sbin/httpd


This will return you the pid of the latest httpd process.

--ahamed

Sure, no problem:
awk '
NR==1 # If it is the first line in the file it is the header, perform the default action, i.e. print the line { print $0 }
...

OK, try something like:
awk 'NR==1; NR>1{A[$2]+=$3; C[$2]++} END{for (i in A) print i,A[i]/C[i]}' OFMT='%.2f' file

Or:
ps -fu myuserid | grep [j]ava > /dev/null &&
echo 'process is running' ||
echo 'process not running'if ps -fu myuserid | grep [j]ava > /dev/null; then
echo 'process is running'
else...

if [ $(ps -fu myuserid | grep -c [j]ava) -gt 0 ]
then echo 'process is running'
else echo 'process not running'
fi

x=$(ps -fu myuserid | grep [j]ava | awk '{print $2}')
if [ ! -z $x ]
then
echo process is running
else
echo process not running
fi

If should have space between brackets...
can reduce one grep
and avoid use of back tick's

try

x=$(ps -fu myuserid| grep -v grep | awk '/java/{print $2}')

if [ $x -eq 0 ]

Should be $13=="" not ! $13; the latter treats a 0 like a "".
If you want to include the error in the output file, you can do that in awk
awk -F, '
($13=="") {print "Error in file"; exit 1}...

Here is another way to do it assuming that you always have 4 input lines / output line and that the input lines are always in the same relative order. It uses slightly different headings than you...

How about this:

awk '
{
col=$0
line=$0
gsub(/ *= .*/, "", col)
gsub(/^[^=]*= /, "", line)
if (!(col in COLS)) {
head=head OFS col
TITLE[++cols]=col
...

How about this
for i in `ls [0-9]-*.txt`
do
mv "$i" "0${i}"
done

for file in *.txt
do
[[ ${file%%-*} -ge 10 ]] && continue
echo "mv $file 0$file"
done


--ahamed

Like I said before, the code was based on some assumptions.

I assumed your file names are abcd1.txt, abcd2.txt ... abcd10.txt

Here is the modified code that might work for you:
#!/bin/bash
...

awk '{ for (i=1; i <=NF; i++) if(substr($i,1,1) == "V") print $i }' file_name


Cheers,

Why don't you try it out for yourself with something like:
for BLOCK in /sys/block/myblock[a-z]*
do
echo "$BLOCK"
done

The 'for' loop can do what you want.

for BLOCK in /sys/block/emcpow*
do
echo "100000" > "$BLOCK"/queue/nr_requests
echo "noop" > "$BLOCK"/queue/scheduler
done

#3 can be...

find /abcd -mtime -1 -type f -exec cp {} /xyz/ \;