viernes, 26 de octubre de 2018

VSAN - Políticas de almacenamiento. Demostración en video.





En este video podemos ver como utilizar las políticas de almacenamiento en VSAN de forma dinámica y cambiar el almacenamiento de la VM sin apagarla.

Podemos ver el tiempo que necesita para arrancar la VM, como creamos después una política que limita las IOPSy la VM tiene una respuesta mucho más lenta al arrancar y finalmente como al cambiar de nuevo la política con la VM en marcha vuelve a tner buen rendimiento.








viernes, 29 de noviembre de 2013

EMC VNX 5200 VIA initialization fails writing sysadmin password

We just initialized a new VNX 5200 and after finishing we couldn't log in with the new sysadmin password we choose. However we could log as root and nasadmin.

So we created a new global user  called temp from ssh connection to the CS and accessed unisphere with the new user and changed sysadmin password.

Below  you can see the command we used:

navicli -h spa security -adduser -user temp -password temp -scope 0 -role administrator

Finally we could enter Unisphere with sysadmin and finish the installation.

jueves, 26 de septiembre de 2013

Error installing Exchange 2013 CU2 on Exchange 2007 organization

Yesterday I installed for first time a new Exchange 2013 on a company with Exchange 2007. We have installed some new 2013 Exchange but this was the first migration.

After accomplish all prerequisites on Microsoft guides we started to install and it ended on error.

The error was "length of the property is too long the maximum length is 64" and it was close to a line referring to Offline Address Book.

In Spanish the OAB was named "Libreta de direcciones sin conexión de exchange".

I googled a little and couldn't find any reference to this error and finally just tried something.

I just renamed the OAB to "OAB" and tried again to install.

This time installation finished OK!!!


lunes, 1 de julio de 2013

EMC File Replication VNX to VNXe

Recently we needed to configure File replication from a VNX array to a VNXe.
I searched EMC support and communities and it was a surprise that there was no documentation or guide.

I have to look for the compatibility matrix again to see if it was possible.

https://elabnavigator.emc.com/vault/pdf/EMC_VNXe_ESSM.pdf

Then I tried to do the inverse  one, VNXe to VNX as I had a manual.

https://www.ndm.net/emcvnx/pdf/vnx/h8270-replicating-vnxe-cifs-nfs-shared-folders-vnx.pdf

At first time I had a problem establishing the trust  and finally it result I needed to add a route at DataMover level on the VNX since the two arrays are in different subnets.

Once I got this solved I followed instructions and it worked fine. Also I tried to make a switch-over on replication and also worked.

Now I created a Replication Destination on Shared Folders screen of the VNXe. The wizard asked which one was the origin and created a Shared folder mounted as only read and with the right size and protection size.

When I tried to set a new replication from VNX as I entered the VNXe as destination the Unisphere screen failed.



I then tried to look for info about doing it on commands from Control Station and I found a blog where some one had the same problem. Thanks to Jon Austin by the way.

http://www.integrateddatastorage.com/blog/2011/11/04/fun-with-darts/

The main problem was how to know the destination FS ID on the vnxe system.

So I connected to ssh on both arrays.

On vnxe I logged as service and entered the service password which was the same that admin one.
I executed  svc_service_shell so I could use commands on the VNXe.
I don't know the password for that since some one form  support put it the day before for other thing and there is a time when it expires. It looks like it takes a couple of days until it expires.

Then I executed /nas/bin/nas_fs -list  so I could see the ID for the replication FS.

After I logged on  VNX CS with ssh as nasadmin user

I executed  nas_cel -interconnect -list so I could see the interconnect id (VNXe SPA)  where the destination FS was published.

and  finally executed this command and finally got replication started!!!

 nas_replicate -create FS_Name -source -fs id=13 -destination -fs id=42 -interconnect id=20004  -source_interface ip=x.x.x.x -destination_interface ip=x.x.x.x -max_time_out_of_sync 10 -overwrite

Overwrite option was needed because it fails other way.

That was all, after 20 hours file system is replicated and working fine.




lunes, 15 de octubre de 2012

Error 3221684229 Installing Exchange 2007 (Mailbox Role) - On Windows Server 2008 R2 and OWA not working

Hi,

this post is composed by a couple of issues that can appear when migrating exchange 2003 to 2007.
Since Exchange 2007 is an old software if you install it today on a Windows 2008 R2 Server you can get this error: Error 3221684229 Installing Exchange 2007 Mailbox Role.

I found this post that helped me out.

http://www.petenetlive.com/KB/Article/0000262.htm

Solution as easy as running the installer on Vista  compatibility mode.

This generated next problem, it was a typical install and all roles where installed except the mailbox.
After changing compatibility mode I added Mailbox Role.

Then we starting trying the server and OWA web page appeared but it didn't authenticate.
Finally I found two posts that help me out:

http://exchangeserverpro.com/modifying-exchange-server-2007-roles-breaks-outlook-web-access

http://itproafrica.com/technology/exchange/owa-broken-on-exchange-2007-on-windows-2008-red-x-appear-instead-of-web-pages/

We needed to install Exchange SP3 and review IIS prerequisites.


lunes, 30 de julio de 2012

VMWARE VIEW Problem cloning master image for linked clones on v5.1

On a small new installation of Vmware View 5.1 we had problems when creating linked clone pools.
It is a lab install on SATA disk and cloning takes about 70 minutes.
It always finished with an error "Connection Server - View Composer Fault: VC operation exceeded the task timeout limit of 0 mins set by the View Composer for <VM instance>"
So after googling a lite bit I  founf some post on Vmware communities   that solved the problem.

http://communities.vmware.com/message/2063937

Thanks to hettefagh I could solve my problem.

Here you can read the solution:


1. Stopped the View Composer service.

2. Made a backup copy of C:\Program Files (x86)\VMware\VMware View Composer\SviWebService.exe.config and saved in the same folder.

3. Edited the SviWebService.exe.config and changed the line

<VirtualCenter PropertyCollectorTimeout="60" />

to read

<VirtualCenter PropertyCollectorTimeout="60" VcTaskTimeout="90" />

4. Saved the change and restarted Composer.  Confirmed the service is running again without error.

miércoles, 7 de marzo de 2012

Esxi5 disconnected and responding ping.


Recently we have some serious problems on a small project implementation.


We had 2 Cisco UCS C200 M2 servers, 2 Cisco MDS9148 and a EMC VNX5100 storage array.
The servers used Qlogic 8GB Fiber channel cards. The problem appeared always on heavy storage use (backup tasks mostly), but it could be once a week.

When accessing to the esxi console you could write user and password but you could not enter.


After that you needed to restart the Esxi and Virtual Machines were down.

I opened a SR on VMware but solved the problem before they answer.

After looking at vmkernel.log I searched on Google for info about errors appearing and finally I founded a KB Article that was very interesting:

vHBAs and other PCI devices may stop responding in ESX/ESXi 4.1 and ESXi 5.0 when using Interrupt Remapping


Also I founded at Vmware  communities   some  threads with same case:



We execute the actions in the KB and everything is working fine after 20 days.


Hope this helps others on searching solutions!!