| |
| |
|
| |
| anewby |
2010-04-29 09:30:35 |
Best available tool - but it would be nice to have more
As flyzone pointed out, the ksh93 version shipped in RHEL5 and OEL5 breaks roracle.
Even with flyzone's mods, you still can't get database checks working because the local DB_USER and DB_PASSWORD variables go out of scope when the database checks are invoked. All you get is a message saying the database is down or hung, even though the previous health check confirms that it's up.
From what I could tell, it would take a pretty significant edit to modify the script to be ksh93-compatible.
As an alternative, I was able to prove that the public domain ksh88 implementation (pdksh) could run the script properly. Here;s what you do:
1. Download and install the pdksh package (yum install pdksh)
2. Replace the first line of bb-roracle.ksh with #!/bin/pdksh
DO NOT use the Linux alternatives command to replace ksh93 with pdksh as the default ksh. Lots of stuff will break.
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
 |
| maquinte |
2007-12-05 05:51:49 |
Nice Tool
Nice tool, easy install, work in RHEL4 u4 with 9i, 10g &RAC
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| flyzone |
2007-11-26 21:09:49 |
not working on RHEL5
make attention on new redhat enterprise server, the ksh shell is changed, the "local" function doesn't work, the script need some change; converting to bash will have however problem: set -a work just on ksh.
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| junieh |
2007-10-29 11:42:12 |
Privileges to the tests
Just to remember, on the script comments it says: only connect and select any table privilege, but it´s necessary to grant a select any dictionary to some tests run correctly, like tablespace check, because a query against dictionary views are performed on the script. So, just to remember this detail... :D
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| epok |
2007-03-05 23:29:59 |
no comment
no comment
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| flyzone |
2007-02-22 22:07:43 |
too many bugs!!
The BB oracle global status it's going wrong. Many users could have a problem with ksh. Much better to substitute line 1834 with:
$GREP "${1}$" ${2} > ${2}1
{ while read input ; do
and the line 2134 "done" with:
done } < "${2}1"
then to add a rm at the last line before the exit:
$RM -f $SIDFILE"1" >/dev/null 2>&1
I have found also a "division by zero for V_TOTUSR and V_USRFAI near line 1758. Must be write a check there.
And about deadlock, i have rewrite the function to put in a table and doing a "if (( $VAR0 % 2 )) then echo '<tr>' fi" to have a doubl-table!!
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| panzam |
2005-09-09 23:26:15 |
Looks Very good
My oradb doesn't appears. BBOUT shows /tmp/SIDS.$$ doesn't exists.
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| ksebesta |
2005-08-30 01:44:58 |
/dev/null bug repaired.
Very configureable.
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
| oboyaval |
2005-08-19 00:20:42 |
big bug in this version
I found a big bug in this script. it erase the device /dev/null on the server. The rm commands are in line 439 and 496. In this two line, the > sign has been forgotten before the /dev/null
|
| |
1 2 3 4 5 |
| Features |
|
| Ease |
|
| Overall |
|
|
|
|
|
|
