ESXi

to find if ESXi is in Physcial or Virtual RDM

for id in ` vim-cmd vmsvc/getallvms |awk '{print $1}'`
> do
> echo $id
> vim-cmd vmsvc/device.getdevices $id|egrep -i "vmdk|compat"
> done
Vmid
(vim.fault.NotFound) {
dynamicType = ,
faultCause = (vmodl.MethodFault) null,
msg = "Unable to find a VM corresponding to "Vmid"",
}
112
fileName = "[NTNX-1-local-ds-NTNX] ServiceVM-1.12_Ubuntu/ServiceVM-1.12_Ubuntu.vmdk",
fileName = "[NTNX-1-local-ds-NTNX] rdms/t10.ATA_____ST91000640NS________________________________________9XG056HW.vmdk",
compatibilityMode = "physicalMode",
fileName = "[NTNX-1-local-ds-NTNX] rdms/t10.ATA_____ST91000640NS________________________________________9XG05B8D.vmdk",
compatibilityMode = "physicalMode",
fileName = "[NTNX-1-local-ds-NTNX] rdms/t10.ATA_____ST91000640NS________________________________________9XG068F8.vmdk",
compatibilityMode = "physicalMode",
fileName = "[NTNX-1-local-ds-NTNX] rdms/t10.ATA_____ST91000640NS________________________________________9XG066HX.vmdk",
compatibilityMode = "physicalMode",
fileName = "[NTNX-1-local-ds-NTNX] rdms/t10.ATA_____ST91000640NS________________________________________9XG04NQP.vmdk",
compatibilityMode = "physicalMode",
128
fileName = "[NTNX-1-local-ds-NTNX] NTNX-vCMA/NTNX-vCMA.vmdk",
192
fileName = "[NTNX-1-local-ds-NTNX] vMA_Customized_Gold_Image/vMA_Customized.vmdk",
fileName = "[NTNX-1-local-ds-NTNX] vMA_Customized_Gold_Image/vMA_Customized_Gold_Image.vmdk",
208
fileName = "[Nutanix-Datastore-1] PDS-PGP/PDS-PGP.vmdk",
240
fileName = "[Nutanix-Datastore-1] pds-wdm1/pds-wdm1.vmdk",
304
fileName = "[Nutanix-Datastore-1] PDS-WebTest/PDS-WebTest.vmdk",
336
fileName = "[Nutanix-Datastore-1] PDS-MSX6/PDS-MSX6.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-MSX6/PDS-MSX6_1.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-MSX6/PDS-MSX6_2.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-MSX6/PDS-MSX6_3.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-MSX6/PDS-MSX6_4.vmdk",
compatibilityMode = "virtualMode",
400
fileName = "[Nutanix-Datastore-1] PDS-SQL7/PDS-SQL7.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-SQL7/PDS-SQL7_1.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-SQL7/PDS-SQL7_2.vmdk",
compatibilityMode = "virtualMode",
416
fileName = "[Nutanix-Datastore-1] PDS-DC11/PDS-DC11.vmdk",
compatibilityMode = "virtualMode",
fileName = "[Nutanix-Datastore-1] PDS-DC11/PDS-DC11_1.vmdk",
compatibilityMode = "virtualMode",
80
fileName = "[NTNX-1-local-ds-NTNX] W2k8_vCenter/w2k8_vcenter_1.vmdk"

FM Troubleshooting

1. When FM fails with PostGres with Security error,

If the install still fails please:

Make sure to install Postgres on the CONSOLE of the server. If you are using
RDP (remote
desktop) for the installation, make sure to start it like this:
Start-->Run-->mstsc
/v: /console

If the installer fails:

* Manually create a LOCAL "postgres" user account (not domain account)
* Make sure the password is a STRONG Password (upper/lower case, digits,
etc)
* Assign the user to the "Power Users" or "Users" group
* DO NOT add the user to the local or domain Administrators group
* Under Local security policy > grant user rights assignment:
o Log on as service
o Log on locally


Once the user has been added, do the following:

* Make sure the postgres user can get a command prompt
o As the LOCAL Administrator, get a command window
(Start-->Run-->CMD)
o In the command shell, type "runas /user:postgres cmd" - this should bring
up a new
command window as the postgres user.
o If this fails, adjust the security policy (via the "cacls" command) to
allow the user to
run the command window (specifically, "cmd.exe")
o Example (in a command shell as the Administrator):

CD \WINDOWS
CACLS *.EXE /E /G "AUTHENTICATED USERS":R
CD SYSTEM32
CACLS *.EXE /E /G "AUTHENTICATED USERS":R


* Make sure the postgres user can write to "NUL" (taken from here:
http://archives.postgresql.org/pgsql-hackers/2006-03/msg00111.php)
o In the command shell (as the Postgres user), type, "dir > NUL"
o If this fails, download and install the "subinacl" tool from Microsoft
(see below for
download location)
o As the Administrator, get a command prompt and type the command, "Subinacl
/service NULL
/grant="authenticated users"=QSEILU"
o Retry the "dir > NUL" command again - it should succeed


If the installer still fails

* Make sure the postgres user can access the files and directories listed in
the section
below (Directories and Files)
* Disable any/all domain policies for the postgres user
* Take the server out of the domain and use local authentication only
* Download the "filemon" utility to identify which file(s) is/are
causing the issues

Filemon - http://technet.microsoft.com/en-us/sysinternals/bb896642.aspx
Subinacl -
http://www.microsoft.com/downloads/details.aspx?FamilyID=E8BA3E56-D8FE-4A91-
93CF-ED6985E3927B&displaylang=en

2. If it is new install make sure you unistall any postgres installation.

-----------------

Things to collect and look for:


- Collect FMs_snmp.log, fmweb.log, Fmserver.log (both PM and FM logs are in there)

- enable snmp trace for the FMS using LogLevel gui; do demand polling

of the fabric;get the fms_snmp.log . ( right click on Log tab and enable snmp

trace), do "circular Arrow on top of FM " to do demand polling and bottom right hand side should say "polling")

- Please do this with Cisco Engineers
In MDS 9000 directory we created debug.properties with allowed=localhost
and did telnet localhost 9197

show fabric
show fabric 2 status
persist 3 ( based on the show fabric)

Cloud 9!

Latest buzzword is Cloud Computing

Utility Computing, Grid Computing, Cluster Computing, Distributed Computing, Provisioning, Application Service Providers, Software as a Service, Hardware as a Service and Software Oriented Architecture

Process Area Networking/Virtualization.
2009 will be the year of Cloud computing! These helps
you improve elasticity of your network/computer/IT
meanwhile having high security.
http://blogs.zdnet.com/BTL/?p=8199

* Deployment speed. One big hang-up for enterprises is figuring out how to procure and provision infrastructure to support a new application. In other words, you can develop an application in two weeks, but wait six weeks to procure and then install the servers that support it. Toss in capacity planning and the time to market expands more.
* Costs. To acquire those additional servers to support a new app requires budget. Staten notes you can’t just run out and buy a server anymore.
* Businesses want fast prototypes. Corporations can deliver faster prototypes by using cloud computing services. Simply put, it makes sense to use cloud computing as a testbed for projects that don’t have a fully-baked business case. For instance, research and development projects, low priority business applications and collaboration services are all good candidates for the cloud.

http://i.zdnet.com/blogs/forr.png

Telepresence.

It is close to OMNIpresence, you are present in every location geographically apart,
if you want to be. Couple of senses are fed - visual and audio. - Visual Networking.

The revolution of phone is taken to next level. With Telepresence booths, which
brings similar to phone booths --- public telepresence .. and also service provider
telepresence. You can reduce travel.

cisco Telepresence

Using the system itself, Cisco says its employees have avoided travel to more than 16,000 meetings, saving the company more than $150 million, and preventing the production of 47,000 metric tons of greenhouse gas emissions.

Procter & Gamble, a Cisco customer, is eyeing similar results. P&G has installed 40 TelePresence systems globally and estimates the infrastructure will eliminate some 1,000 business trips per month. Of course, Cisco’s TelePresence solution is an elite corporate offering, with units priced up to $300,000. However, the company says a SoHo version is in the works, with availability planned in about a year. A long list of competitors is attacking the category as wel

With reducing environment footprint, soon it will be cheaper to host telepresence.

With broadband speed,eventually telepresence will come to each home.. highspeed
network, big TV and good video camera.. Here is Web 3.0.

Video conference Telepresence helps to get visual clues.With travel cuts, this
seems good option. Cisco's growth was 5 times...

MDS Fc port stuck in initializing

Reasons:
- flogi server not responding correctly
show flogi internal event interface fc x/y

- persistent fcid limit exceeded ( 255 entries with all area unique fcid because
of qlogic hbas)

MDS9513-83-SJ# show run | include "vsan Z" | include fcid | count
where Z is vsan #.
- conflict in persistent fcids ( MDS can't assign the fcid allocated
to WWN)

- # of devices more than 255 if qlogic or devices that need area unique fcid
is allocated

- supervisor failure or module issue, flogi command not reaching the supervisor
or flogi process.

show fc2 internal even errors

these are few causes.

Flex Attach in MDS!

MDS gives virtual pwwn that is used for zoning and lun masking, so when a host
connected a port dies, you can either connect a new host to same port or reconfigure
flexattach config on port where spare host is connected. So this is a security concern
because anyone can come and attach another host to port and get all the access to luns.
Similar way someone can replace or remove hba and connect it to a different server.

This can be reduced by using port-security.

http://www.cisco.com/en/US/docs/storage/san_switches/mds9000/sw/rel_3_x/command/reference/CR03_f.html#wp1393061

MDS callhome for bootflash errors!

When a bootflash fails, if callhome is configured correctly, we would expect
XML destination to send the callhome, it does not happen even if Cisco_TAC alert group
is added because of the bug
CSCso71302.
Workaround is defined in the above Cisco MDS bug.

Symptom:

BOOTFLASH failure does not generate callhome to default XML destination profile.

Workaround:
Default XML destination profile is configured to receive alerts from Cisco-Tac alert group by default. You need to add linecard-hardware and supervisor-hardware alert groups to that destination profile to get callhome messages like BOOTFLASH failure


for eg.
callhome
destination-profile xml message-level 2
destination-profile xml email-addr auto-notify@cisco.com
destination-profile xml alert-group linecard-hardware
destination-profile xml alert-group supervisor-hardware
destination-profile xml alert syslog-group-port