KSar: Generating Graphs from SAR Reports

We all know the SAR (System Activity Report), however sometimes it’s dificult to visualize a large amount of data or even extract some long term meaningful information.
How wonderful would be to have a graphical visualization from this data? Well, it’s pretty simple using KSAR.

KSAR is a BSD licensed Java based application to create graph of all parameters from the data collected by Unix sar utilities and can be exported to PDF, JPG, PNG, CSV, TXT and others.
The project Codes are here. The latest Version is KSar2-0.0.4.

See below an I/O Graph from month of Dec, generated from a database server, as an example:

GrepOra-srv.jpg

To use it, first thing is to have SAR data. To get it we have basically 3 options:
A. Collect from current server.
B. Extract from other server using direct SSH connection.
C. Use a Generated SAR File
D. Run Java tool from Client Server.

Personally, I prefer to use option C, in order to avoid putting any code in client servers and also work in less intrusive mode as possible.
I also don’t use option B because we don’t usually have direct connection to client server, but sometimes with jumpboxes or similar.
There is a third reason: When Chosing option A or B, it’s automatically connected only daily data, but when using C, you can put all data you need. It need only to be available on server.

For reference regarding Option D, please check this link.

By the way, some other useful information about SAR:
1. SAR Collection Jobs can be checked on /etc/cron.d/sysstat
2. SAR Retention can be checked/adjusted on /etc/sysconfig/sysstat

Ok, now how to generate the SAR Files?
Using command: sar -A

Example:

[root@grepora-srvr ~]# cd /var/log/sa/
[root@grepora-srvr sa]# ls -lrt |tail -10
total 207080
-rw-r--r-- 1 root root 3337236 Dec 24 23:50 sa24
-rw-r--r-- 1 root root 3756100 Dec 24 23:53 sar24
-rw-r--r-- 1 root root 3337236 Dec 25 23:50 sa25
-rw-r--r-- 1 root root 3756113 Dec 25 23:53 sar25
-rw-r--r-- 1 root root 3337236 Dec 26 23:50 sa26
-rw-r--r-- 1 root root 3756104 Dec 26 23:53 sar26
-rw-r--r-- 1 root root 3337236 Dec 27 23:50 sa27
-rw-r--r-- 1 root root 3756096 Dec 27 23:53 sar27
-rw-r--r-- 1 root root 3337236 Dec 28 23:50 sa28
-rw-r--r-- 1 root root 3756100 Dec 28 23:53 sar28
-rw-r--r-- 1 root root 2317668 Dec 29 16:30 sa29
[root@grepora-srvr sa]# sar -A -f sa29 > sa29.txt
[root@grepora-srvr sa]# cat sa29.txt |head -10
Linux 3.8.13-118.4.2.el6uek.x86_64 (grepora-srvr) 12/29/2017 _x86_64_ (40 CPU)
12:00:01 AM CPU %usr %nice %sys %iowait %steal %irq %soft %guest %idle
12:10:01 AM all 97.74 0.00 1.71 0.01 0.00 0.00 0.52 0.00 0.02
12:10:01 AM 0 96.46 0.00 2.59 0.02 0.00 0.00 0.92 0.00 0.01
12:10:01 AM 1 98.55 0.00 1.24 0.01 0.00 0.00 0.20 0.00 0.00
12:10:01 AM 2 97.83 0.00 2.04 0.01 0.00 0.00 0.11 0.00 0.02
12:10:01 AM 3 98.44 0.00 1.41 0.01 0.00 0.00 0.14 0.00 0.01
12:10:01 AM 4 98.28 0.00 1.65 0.00 0.00 0.00 0.06 0.00 0.01
12:10:01 AM 5 98.27 0.00 1.70 0.00 0.00 0.00 0.02 0.00 0.00
[root@grepora-srvr sa]#

With this file you can copy it from client server your server and import using KSAR Interface. It’s pretty intuitive and easy to use.

But how to generate all available days or a set of specific days in past?
Here is a script I use for this:

### All Days of SAR
DT=$(ls /var/log/sa/sa[0-9][0-9] | tr '\n' ' ' | sed 's/\/var\/log\/sa\/sa/ /g')
## Explicit Days
#DT="07 08 09"
#DT="12"
# Today
#DT=`date +"%d"`
>/tmp/sar-$(hostname)-multiple.txt
for i in $DT; do
LC_ALL=C sar -A -f /var/log/sa/sa$i >> /tmp/sar-$(hostname)-multiple.txt
done
ls -l /tmp/sar-$(hostname)-multiple.txt

After this you can copy the generated file to you PC and generate the same report.

Hope you enjoy it!

Cheers!
Matheus.

Advertisements

Finding Trace Files Being Written Right Now!

Hey!
I was not sure on the title for this post, but I bet everyone, at least once, needed to know which file is being modified at this exact moment in your filesystem/server.

Some days ago I noticed something was making my filesystem full. I cleared some gigas in logs from Diag Home but the space gone 100% very quickly. What is consuming the space?
Easy:

1. Create a new file.

$ touch a.log

2. Find everything newer than this file.

$ find . -newer a.log

Here you go!

In my situation, after finding this, I noticed there was a session in a bug situation generating thousands of messages on trace file.
Killed the session, got part of the messages, cleared file. Issue solved.

Hope it helps!

CRS Not Starting after Removing OS User: How to Workaround and How to Solve!

Hello all!
Turns that a few days ago a client reached me because his CRSD was simply not starting. Like this:

[root@proddb proddb]$ ./crsctl start res ora.crsd -init
CRS-2672: Attempting to start 'ora.crsd' on 'proddb'
CRS-2676: Start of 'ora.crsd' on 'proddb' succeeded

[root@proddb proddb]$ ps -ef |grep crsd
root 19217 13424 0 11:53 pts/0 00:00:00 grep crsd

After some investigation, I found the following:

2017-01-24 14:00:06.859: [ CRSSEC][1690195712]{1:51052:2} Exception: OwnerEntry construction failed to retrieve user id by name with ACL string: owner:jacknobody:rwx and error: 1
2017-01-24 14:00:06.912: [ CRSSEC][1690195712]{1:51052:2} Exception: ACL entry creation failed for: owner:jacknobody:rwx

Hmmm, seems some CRS resources are owned by “Jack Nobody”… Turns that I this us was removed from OS:

[root@proddb proddb]$ cat /etc/passwd |grep jacknobody
[root@proddb proddb]$ 

What to do now?

Continue reading

umount: This utility only unmounts cifs filesystems

Can’t umount Red-Hat filesystem with error “This utility only unmounts cifs filesystems” either with force?

[@Linux-redhat5 ~]$ cat /etc/redhat-release
Red Hat Enterprise Linux Server release 5.11 (Tikanga)

ERROR:

[@Linux-redhat5 /]# umount /cifsfilesystem/data/custom
This utility only unmounts cifs filesystems.
This utility only unmounts cifs filesystems.
This utility only unmounts cifs filesystems.

Continue reading

Shellscript: Which database is this service_name from?

Hey!
Strange question, right? You should know you applications and databases and how everything works, right?

Happen that some days ago a client requested me to identify from which database is a service_name. Easy? Not if the server have 46 database instances running…

Ok, now how can I can this information as quick as possible?

I know that the command below give me all distinct databases in CRS:

/u01/app/12.1.0.2/grid/bin/crsctl stat res -t -w "TYPE = ora.database.type"

Now, lets make some sed magic:

[root@nprd01dbadm01 ~]# /u01/app/12.1.0.2/grid/bin/crsctl stat res -t -w "TYPE = ora.database.type" |grep ora. |sed 's/ora./srvctl status service -d /' |sed 's/.db/ |grep \$1/' | head -1
srvctl status service -d dba |grep $1

Continue reading

Teradata ODBC for Unix OS

To connection between unix machine with teradata environment is needed to install ODBC Driver and some others components (dependencies). All Items can be found in Teradata Downloads site: https://downloads.teradata.com/

Teradata OBDC Driver has 3 dependecies:

1. Shared common components for Internationalization (TDICU)
2. Teradata GSS client package (TeraGSS)
3. Teradata Tools and Utilities Base (TTU)

After download all items, run as root: rpm -ivh “component_name” –nodeps

Eg: rpm -ivh teragss1510-15.10.02.06-1.noarch.rpm --nodeps

Use “–nodeps” only to not need to install in order

Continue reading