My colleague has set up log ingestion with GreyLog quite successfully. I
keep trying to persuade him to speak at a conference on the subject.
Patrick
On Wed, 16 Dec 2020 at 16:35, Sayan Malakshinov <xt.and.r@xxxxxxxxx> wrote:
Hi Nenad,
We use V$DIAG_ALERT_EXT in Zabbix. Previously we used x$dbgalertext, then we
used some tricks like `with pre as (select /*+ materialize */` when we
started using V$DIAG_ALERT_EXT to avoid different bugs in it.
On Wed, Dec 16, 2020 at 7:18 PM Noveljic Nenad <
nenad.noveljic@xxxxxxxxxxxx> wrote:
This is the question for those who monitor alert log with their own
script. (We don't like OEM.)
How do you monitor it: by querying V$DIAG_ALERT_EXT or parsing the adrci
output?
Do you see any clear advantages or pitfalls with one or the other?
We wrote a script based on adrci 10 years ago, which has worked well.
Then, we instantly eliminated the V$DIAG_ALERT_EXT approach because older
database versions where throwing ORA-00600. Now, as we have to rewrite the
script, we would like to evaluate both options again. By the way, the
reason for rewriting is the transition from Perl@Solaris to Python@Linux.
Thanks,
Nenad
--
Best regards,
Sayan Malakshinov
Oracle performance tuning engineer
Oracle ACE Associate
http://orasql.org
