Bert Zefat's Tech Blog

In one instance we encountered following error on a VEEAM backup environment, where all backup’s failed in the same session. also manually restarting the jobs gave this error:

Unable to subscribe to guest processing components: Retrieving the COM class factory for component with CLSID {D2E3C73F-841A-46B0-A93F-956CCABD3B11} failed due to the following error: 80080005 Server execution failed (Exception from HRESULT: 0x80080005 (CO_E_SERVER_EXEC_FAILURE))

After investigation it seemed that intl.dll was missing in the …\Veeam\Backup and Replication\Backup\VSS folder.

Replacing it fixed the problem.

Filed Under: VEEAM Add Comment »

OK. This week for serious. Going for App-v. After first doing the necessary (Pre)-Intakes, this week finally going for Packaging.

Nice for a Linux guy, doing Windows ‘debugging’.

However. A very nice step for me and still wishing to learn more. Tomorrow an other day.

App-V …. Here I come :-)

Filed Under: App-V, Microsoft, Windows Add Comment »

After upgrading my laptop and the last reboot I found out my Start button (of my first (laptop) screen) did not react to my mouse clicks. (On my second screen it worked fine).

Seems you should do next:

1 – Go to: C:\Users\Default\AppData\Local\Microsoft\Windows and copy the WinX folder.
2 – Go to: C:\Users\[your user name]\AppData\Local\Microsoft\Windows and replace the WinX folder

(I zipped the old one, just to be sure)

Log out and logon again.

Start button works fine again

Have fun!

Filed Under: Windows Add Comment »

Once in your lifetime you encounter that you have changed something in one SAN switch. And after a while you see that the other switch(es) where not connected to the switch where you made zoning changes.

If you join them again (fixed it by replacing an defective SFP that was the up-link to my changed switch) you’ll see in the CLI on the ISL port when you enter ‘switchshow’ that the ISL port gives “Segmented”.

To fix it (nicely) you must change both switches in a stats that both have OR the same config again OR you just clear the zoning on one switch (as it should be redundant).

BUT in most environments it’s not applicable  so:

All done!

Good luck.

Filed Under: General Add Comment »

Filed Under: General Add Comment »

While re-starting a ESXi 5.1 cluster. It came back, but one host did not join the cluster.

Adding it to the cluster (or at fisrt during a “connect”) it gave me the following error:

vmodl.fauly.hostcommunication 

Removing and adding it again to the cluster did not work. Disabling and Enabling HA did not do the trick.

Fix:
As it was ESXI 5.1 you connect to it (console or SSH) and type

services.sh restart

Now you can add the node again to the Cluster (remember to Enable HA again)

Filed Under: ESX, VMware Add Comment »

I had to update an MediaWiki server due to security issues:

Things I had to do to get it working:

Added in /etc/my.cnf
net_buffer_length=1000000
max_allowed_packet=1000000000

To make import possible from database dump.

Placed new mediawiki server next to the old one. Copies the images en plugins.

Run from the commandline php update.php

All went well.

Filed Under: Linux Add Comment »

Looking for the best practice regarding the amount of managers that should be running in a Single Site config. I saw that this was not that easy to find (as HP is aiming on Multi-Site Clusters).

For a single site next does count.

2 Nodes, 1 FOM, 3 managers
3 Nodes, 0 FOM, 3 managers
4 Nodes, 1 FOM, 3 or 5 managers (5 preferred)

Just make sure you have an odd manager count, so 1 node can fail ad you still will have 2 managers running.
This could mean that you could run without a FOM in a 4 Node SS Cluster. (3 managers running on 3 Nodes, but nothing may go wrong when 1 node already has failed.)

Filed Under: P4000, Storageworks Add Comment »

Upload the HPPP_x.xx.xx_Linux_X64.rpm to the vMA (/tmp) (with winSCP of FASTscp)

> sudo -s
> rpm -i hppp-linux-1.02.028-1.x86_64.rpm

Starting Power Protector: HP-HPPP[ OK ]

Check if the service is indeed listening on it’s ports:
> netstat -tanp | grep LISTEN
tcp 0 0 0.0.0.0:4679 0.0.0.0:* LISTEN 7813/HP-HPPP
tcp 0 0 0.0.0.0:4680 0.0.0.0:* LISTEN 7813/HP-HPPP
tcp 0 0 :::22 :::* LISTEN 2082/sshd

Default the vMA does not have ports 4679 (http) en 4680 (https) open

Configure communication between HPPP and the HP Network Module:
> sudo -s
> iptables -I OUTPUT -p tcp –dport 80 -j ACCEPT
> iptables -I INPUT -p tcp –dport 4679 -j ACCEPT
> iptables -I INPUT -p tcp –dport 4680 -j ACCEPT
> iptables -I INPUT -p udp –dport 4679 -j ACCEPT
> iptables -I INPUT -p udp –dport 4680 -j ACCEPT
> iptables -I OUTPUT -p udp –dport 4679 -j ACCEPT
> iptables -I OUTPUT -p udp –dport 4680 -j ACCEPT

> service iptables save
> iptables-save

Now you can connect with a web-browser.
Also the commi=unication between the vMA HPPP and the UPS is restored.

Filed Under: ESX, VMware, vSphere Add Comment »

If the issue persists, end any vmware-vmrc.exe processes in the Windows Task Manager:

1. Close all vSphere Client sessions.
2. Open Windows Task Manager.
3. Search for any vmware-vmrc.exe processes and end the process.
4. Start the vSphere Client and connect to the host directly or to vCenter Server.

Filed Under: ESX, VMware, vSphere Add Comment »