Increasing the size of virtual SCSI disks on the Virtual I/O Server

Today, I had a request from our client to extend a filesystem. The problem, as usual, is that there was no space available in the VG, and there weren't disks available in the system, so I had to add one new disk.

The interesting thing is the environment, this system doesn't receive disks  from a storage cabin, the disks are asssigned from on VIO server.

In this case, it is one virtual SCSI disk that it is backed up by a logical volume.

I thought that I had to create a new virtual disk for the client machine but I found out that it can be done expanding the size of the existing disk.

So these were the steps that I took 

• Extend the logical volume in the first VIOs (I have two VIOs serving the disk an there is a mirror between both disks at the client level)

# extendlv dbdes_saplv 20G
0516-622 extendlv: Warning, cannot write lv control block data.
0516-622 extendlv: Warning, cannot write lv control block data.

I got this error but as you can see in the second link in the reference section, you don't have to worry about.

• Extend the logical volume in the second VIOs  

# extendlv dbdes_saplv 20G
0516-622 extendlv: Warning, cannot write lv control block data.
0516-622 extendlv: Warning, cannot write lv control block data.

•  On the client, adjust disks to the new size

# bootinfo -s hdisk2 
# bootinfo -s hdisk3

• You have to make the VG be aware of the growth of disks

 chvg -g sapvg



Note: Maybe the size of the disks take a little bit showing the new size but if you run lsvg VGname, you will see the new free space in the Volume Group

References

http://www-01.ibm.com/support/docview.wss?uid=isg3T1011227

http://www-01.ibm.com/support/docview.wss?uid=isg1IV35658

Configure NTP in AIX

I will start by defining Network Time Protocol (NTP) which is a networking protocol for clock synchronization between computer systems.

To configure it, you have to edit the following file /etc/ntp.conf where you will configure the ntp server 

server ntp.time.es

Leave the driftfile and tracefile at their defaults. 


If you want that xntpd starts when the system boots, you have to uncomment the following line in /etc/rc.tcpip

# Start up Network Time Protocol (NTP) daemon
#start /usr/sbin/xntpd "$src_running"

To verify that your server is synchronizing properly with your ntp server, restart the daemon by running the commands below

stopsrc -s xntpd 
startsrc -s xntpd

wait for some minutes and check with lssrc -ls xntpd. You should pay special attention to two lines

The sys peer should addres to your server and sys stratum should be a number between 1 and 15 (Note: see the first reference link for more information about this). If this is not in this way, your server is not synchronizing properly

Good configuration:
Sys peer:        ntp.time.es
Sys stratum:     3

Bad configuration:
Sys peer:        no peer, system is insane
Sys stratum:     16

Another way of verifying your ntp server is by running ntpdate

ntpdate -d ntp.time.es

If it is working you should see information about the ntp server and synchronization details.  

If there is something wrong you will get the message "no server suitable for synchronization found"

If this is the case, check your firewall for any rule that could be blocking the communication

You can also synchronize manually the server by running (xntpd daemon has to be stopped)

ntpdate ntp.time.es

References

 

http://en.wikipedia.org/wiki/Network_Time_Protocol

http://www.ntp.org/

http://www-01.ibm.com/support/docview.wss?uid=isg3T1000653

http://aix4admins.blogspot.com.es/2011/09/time-synchronizationtimed-ntpd-setclock.html