HP-UX how to get all of the syslog

Hi,
I want to get all of the syslog,which is under the catalogue of var/adm/syslog,in order to do something about correlation analysis.
But for days,I can not find anything about that,who can help me?
Thank you very much!

syslog.log is a file. It gets rotated out through a sequence of names, if log rotation has been set up (I made these names up because I do not know what your sysadmin did):

syslog.log is the current file
syslog.0  
syslog.1
syslog.....syslog.n

I do not know how your sysadmin set up log rotatation. Plus the sysadmin can specify that syslog is named something else. If you have no rotation or toher setups the syslog.log file gets overwritten at bootup. And so it is usually really large.

What is your problem, do you get an error? Unless your sysadmin is really paranoid, all of those logs should be readable. Please post the command you used and the output or error.

By default there is no syslog rotation.

Every time you restart syslog (or machine for that matter), syslogd (/sbin/init.d/syslogd) init script will mv the current syslog to OLDsyslog.log in /var/adm/syslog/ directory, and create a new one.

Hope that helps
Regards
Peasant.

@zhaoyy
Please rephrase post #1 in better English. The post does not make sense. Can you post and example?

Thanks for answers. Because of my poor English, maybe I describe not clearly. I want to know where I can get all of the fault information in HP-UX, and the detailed description about this fault information.

Please mention what version of HP-UX you have.
Are you a trained administrator or someone new to HP-UX?

Please give examples of what you want to know. We still do not understand your question.

If you have a support contract, it may be easier to talk to HP in your own language.

Thanks for answer. I want something just like the AIX�s �Error Record Template Repository in HP-UX, so I can find some relationship between the fault information. This is meanings I want to get all of problems write to the var/adm/syslog/syslog.log that the system may be happened. The version of HP-UX is 11i.

The flat file /var/adm/syslog/syslog.log is not suitable for storing every event which can happen on a HP-UX 11.11 system (I assume that is what you mean by 11i) ... or any unix system for that matter.

There are separate detailed but compact databases maintained by the Enviromental Mangement System (EMS) to record all hardware events including the trivia such as correctible memory errors and routine changes of exchangeable media. Any serious hardware errors are also recorded in syslog.log and each message includes the EMS command to get the detail of the error.

I feel that you are presenting us with your non-viable solution to your problem, rather than presenting the problem itself.

I still suggest that you approach HP in your own language with your requirement. They may try to sell you HP Vantagepoint (aka Openview) and some consultancy.

I want to do some fault association analysis in HP-UX. where I can get all of the system fault information, also the detailed description. The most important is the information I want, not just the fault occured on my computer, but all possible fault that the system may be happened, it is just like a repository.
Thank you very much.

Your English is hard to follow.

I think that you are asking for a database of every possible fault which could occur on a HP-UX system.
The hardware and software error monitoring is extensive in HP-UX and this would be a very large database.
I don't believe that this information is available outside of HP.

Please bear in mind that the more serious hardware errors are monitored by the GSP (Guardian Service Processor) and reported to HP-UX where this is possible. The GSP may be actually be called something else depending on what hardware you have.

I also wonder why you want to take this approach. I've never come across a situation where it is necessary. If you do not purchase a commercial monitoring suite of programs, then it is necessary to write your own.

1 Like

Thanks for answer. Your understanding is correct. I just want to get that database. When I know every possible fault, I can do some association analysis in network management field. To do so, can reduce the alarm number.