제2의 비엔지니어 인생관을 꿈꾸며

Filed under Study

If you want to know the SMB version that Windows is using, connect a share, open an admin cmd window and run


Start with current LTS stable: 151014  LTS (long term stable) and optionally update to a newer release
what allows a go back to LTS

Manual Installation:

1. download OmniOS  stable


You can setup OmniOS from USB  (use .dd file, 1 GB+ USB stick) or CD (.iso)
You can use this imager www.napp-it.org/doc/downloads/usb image.zip
to transfer the usb image to your stick (Windows)

2. Install OmniOS
2a. Option: Install on ESXi, create a new VM (Solaris 10-64) with e1000 and/or  vmxnet3 vnic  (16 GB disk min, 25GB+ suggested)
use 2 GB min RAM (best is 8GB+  or 1/3 of available) and 2 cores
Read my manual about setup napp-in-one: http://www.napp-it.org/doc/downloads/napp-in-one.pdf

2b. Install on Hardware:
You can install Omni + napp-it onto Sata or fast USB sticks >=32 GB)
If you use USB, use reliable/fast sticks, best are USB enclosures with an 2,5" Sata SSD

You can also use one or two Sata DOMs for a boot mirror, see http://www.supermicro.com/products/nfo/SATADOM.cfm
(prefer this over USB)

Read my manual about setup napp-it: http://www.napp-it.org/doc/downloads/napp-it.pdf

Reboot now!

3. After installation, login as root (no pw)

4. setup network. Network is set to (per default): network/physical:default,
enter blue commands at console

4.1 create interface
- list available interfaces and use linkname ex e1000g0:

(optional: enable Jumbo Frames first: dladm set-linkprop -p mtu=9000 e1000go)

dladm show-link
ipadm create-if e1000g0

5. use DHCP (or skip to 6.0 to use a static adress)
ipadm create-addr -T dhcp e1000g0/dhcp

5.1 add nameserver
echo 'nameserver' >> /etc/resolv.conf

5.2  use dns (copy over DNS template)
cp /etc/nsswitch.dns /etc/nsswitch.conf

If something happens (typo error), retry, opt. delete interface ex ipadm delete-if e1000g0

5.3 install napp-it (newest free)
wget -O - www.napp-it.org/nappit | perl

5.4 l
start napp-it from your Desktop via Browser
http://serverip:81 (use real server-ip of your server)
You can check your ip at console with command ifconfig

reboot after installation of napp-it  (or check napp-it menu "snapshots - be" that current BE is default)
reboot (mostly not needed)

Next step 6. is an option to 5. (use either 5. dhcp or 6. manual setup)

6.0  add static IP address
  6.1 create static adress
  ipadm create-addr -T static -a e1000g0/v4

  6..2 add default route
  route -p add default

If something happens (typo error), retry, opt. delete interface ex ipadm delete-if e1000g0

6.3 install napp-it (newest dafault for Free/Pro)
wget -O - www.napp-it.org/nappit | perl

6.4 login to napp-it via any web-browser
(you can request ip via ifconfig -a )
http://ip of your server:81

reboot after installation of napp-it  (or check napp-it menu "snapshots - be" that current BE is default)
reboot (mostly not needed)

Modify OS defaults

The OmniOS defaults are suboptimal regarding network settings (ip-buffer), NFS settings, AHCI Hotplug behaviour (disabled per default) or disk timeouts (up to 3 minutes on problems, way to high should be reduced to a TLER alike timeout of 7s).

On napp-it Pro, you can enable improved settings with the base tuning option in System > System Tuning

7. enable remote root access
Napp-it menu Services-SSH allow root

(Re) Enter a root pw at consoler: passwd root (this creates a SMB pw as well)
You are now able to connect via WinSCP (or putty) as root
passwd root

8. opt.  netatalk 3.1
Using the repo http://scott.mathematik.uni-ulm.de/
wget -O - www.napp-it.org/afp | perl 

9. opt. Update OmniOS to newest
pkg image-update (or pkg update)

10.  opt. install TLS email

Check OmniOS repos for other apps:

http://pkg.omniti.com/omniti-ms/en/index.shtml and

or use the pkin repo from Joyent
see http://napp-it.org/downloads/binaries.html

10. Vmware-tools (ESXi 5.5) and OmniOS

If you want to install VMware tools in OmniOS, you can either mount the tools iso manually according to http://napp-it.org/doc/ESXi-OmniOS Installation HOWTO en.pdf within ESXi or

- open the Installer ESXi iso with 7Zip (file VMware-VMvisor-Installer-5.5.0-1331820.x86_64.iso)
- then open TOOLS.T00 and then the file TOOLS with 7zip

- Open the folder 5.5.0/vmtools
You can see all tool-isos now

- Open Solaris ISO
- Open vmware-solaris-tools.tar.gz
- Open vmware-solaris-tools.rar

- Copy the folder vmware-tools-distrib to your OmniOS /root folder (ex use WinSCP)
- change permissions of this folder to 777 recursively
- open a console on OmniOS (as root) and run perl /root/vmware-tools-distrib/vmware-install.pl

Be aware of problems with e1000 vnics on ESXi 5.5 (see http://napp-it.org/downloads/index en.html )
I suggest to use vmxnet3 vnics until this proble is solved by Vmware

Ensure that the Fibre Channel HBA is Detected

# prtconf -D | grep qlc
pci1077,15d, instance #0 (driver name: qlc)
pci1077,15d, instance #1 (driver name: qlc)

Determine the PCI Address of the Fibre Channel HBA

# mdb -k
Loading modules: [ unix genunix specfs dtrace mac cpu.generic uppc pcplusmp scsi_vhci zfs sd ip hook neti arp usba uhci stmf stmf_sbd qlc fctl sockfs md lofs random idm crypto smbsrv nfs sppp ipc ptm fcp cpc fcip nsmb ufs logindmux ]
> ::devbindings -q qlc
ffffff01cb184040 pciex1077,2532, instance #0 (driver name: qlc)
ffffff01cb183d90 pciex1077,2532, instance #1 (driver name: qlc)
> $q

Now we will need this from the above output pciex1077,2532.

Deactivate the QLC Driver for the Fibre Channel HBA

Remove the default driver (qlc) from the adapter.

# update_drv -d -i 'pciex1077,2532' qlc

Activate the QLT Driver for the Fibre Channel HBA

Add the target driver (qlt) to the adapter.

# update_drv -a -i 'pciex1077,2532' qlt

Reboot the Machine

# reboot

Confirm the HBA is using the QLT Driver

# prtconf -D | grep qlt
pci1077,15d, instance #0 (driver name: qlt)
pci1077,15d, instance #1 (driver name: qlt)

# mdb -k
Loading modules: [ unix genunix specfs dtrace mac cpu.generic uppc pcplusmp scsi_vhci zfs sd ip hook neti arp usba uhci stmf stmf_sbd fctl sockfs lofs random idm crypto cpc smbsrv nfs fcip sppp ufs ptm ipc ]
> ::devbindings -q qlc
> ::devbindings -q qlt
ffffff01cb184040 pciex1077,2532, instance #0 (driver name: qlt)
ffffff01cb183d90 pciex1077,2532, instance #1 (driver name: qlt)
> $q

Configure the SCSI Target Service

# svcs stmf
STATE          STIME    FMRI
disabled       13:27:08 svc:/system/stmf:default

# svcadm enable stmf

# svcs stmf
STATE          STIME    FMRI
online         13:29:54 svc:/system/stmf:default

Display the Status of the SCSI Target Service

The ALUA Status is what allows a LUN to be served via FC and iSCSI at the same time.  We do not require or want it.

# stmfadm list-state
Operational Status: online
Config Status     : initialized
ALUA Status       : disabled
ALUA Node         : 0
2016/04/20 21:51 2016/04/20 21:51
Filed under Study
Unable to add esxi host to vcenter Datacenter.QueryConnectionInfo
사용자 삽입 이미지
Cannot contact the specified host. The host may not be available on the network, a network configuration problem may exist, or the management service on this host is not responding
There are lots of actions you need to follow. In my case non of those action resolved my issue except one step no#5
Here are the quick steps you need to follow to find out why you are having Datacenter.QueryConnectionInfo error message while adding host to vcenter.
  1. Lockdown mode is enabled on the ESXi host
  2. Incorrect credentials are entered when adding the ESXi host.
  3. Try full fqdn and add again.
  4. Check DNS settings on vcenter and ESXi host
  5. Your vcenter does not support the ESXi host build number. This issue started to appear after recent upgrade on Esxi 5.5 servers. If your vcenter server is below  vCenter Server 5.5 Update 3b you cannot add host running ESXi 5.5 Update 3b. You need to upgrade your vCenter server to 5.5 U3b in order to add ESXi 5.5 U3b host.
ESXi Build number 3248547
vCenter Server 5.5 Update 3b Build No 3252642

필자는 5번째 이슈에 걸렸습니다. 똑같은 5.5 버젼인데 OS도 바꿔보고 지웠다 깔고 서비스 재시작하고 여러번을 반복해도 원인을 알수가 없었습니다.
구글에 검색해보니 5.5 버젼에서만 Update 빌드버젼이 다르면 호스트 추가 오류가 발생 한다는걸 알게되었습니다. 벰웨어도 잠깐 공부좀 해볼까 해서 설치하는데 이런데서 막히니까 그냥 포기할려다가 해결된 케이스였습니다. Hyper-V 에만 손에 익숙해서인지 보통 벰웨어가  구축 및 설정이 더 편하다고 인식되었는데 그냥 복불복 아닌가 생각이 들고 있습니다.
2016/03/06 22:54 2016/03/06 22:54
Filed under Study
http://social.technet.microsoft.com/wi ··· 016.aspx

OpenGL API support (up to OpenGL 4.4)
  • OpenCL API Support (up to OpenCL 1.1)
  • Expanded the amount of dedicated vRAM assignable to a VM to 1GB
  • Decoupled vRAM from Monitors and Resolution
  • Example applications: Adobe Photoshop CS, ArcGIS Pro, Solidworks, etc.
  • Added support for RemoteFX vGPU when running Server VMs as a VDI guest
  • Added support for Generation 2 VMs
  • Improved end to end user experience
    • Higher Render FPS and lower Jitter
    • Improved application compatibility
일단 위내용데로 지원범위가 넓어졌고 그래픽 가상화 부드러움의 선두주자 젠앞에서는 RemoteFX는 명함도 못내밀 정도였습니다. 특히 캐드 사용자들의 불만이 가장 높았고 일부 프로그램에서는 3D가속지원이 되지 않는다고 기능이 추가되도 전혀 사용도 하지 못하는 사례가 많은 MS 가상화였습니다. 이제까지 서버 가상화에서만 쓸만한 제품이 아니라 VDI 시장에 많은 영향력을 발휘할거란게 필자의 판단입니다. 이제까지 MS가 뭐 지원한다고 떠들어도 막상 써보면 영 아닌데?라는 생각을 한게 한두번이 아니였습니다. 젠은 쓸만은 한데 이기능 저기능 쓸라면 라이센스 비용의 부담이 많고 여러가지 고민을 하게되는 그래픽 가상화의 장벽이였습니다.
그래서 필자가 Windows 10의 RemoteFX 가상화 한대를 올리고 테스트를 해봤습니다.

사용자 삽입 이미지
사용자 삽입 이미지
위그림과 같이 말많았던 오토캐드 성능 테스트와 포토샵까지 사용을 해봤습니다. 캐드가 상당히 부드럽게 돌아가는걸 느낀후 이제는 말할수 있다 RemoteFX편을 만들고 싶었습니다. 포토샵이야 2D를 주로 사용하는 프로그램이라 고만고만 하였지만 그림처럼 하드웨어 가속 활성화가 표기가 되어 있었습니다.

마지막으로 테스트를 Windows 10/Server 2016 이 아닌 Hyper-V Server 2016 Core Free 버젼으로 테스트를 진행할 계획입니다.
2015/12/28 22:00 2015/12/28 22:00
Filed under Study

If you want to be the best at something you gotta start somewhere, get stuck somewhere & find your way out.

Today I am going to show you how to remove Exchange Server from a network when you are unable to uninstall it / You have created a new server & want to get rid of the old server but it wont go away easily.

Word of Warning!

Now I am doing this on VMware Workstation Test Lab and you will have to be very very careful and double check what you are doing before you do it if you are on a production environment , because deleting something from any of these steps will cause serious network & server instability if not done correctly. The recommended way to remove an Exchange server from your network is to properly uninstall it. If the installation no longer exists, then there is also the option of installing a new server with the same name and restoring the original Exchange configuration. However, I have not explored this path as it appears to be much more time consuming.

If you removed an Exchange 2007/2010/2013 server from your network without properly uninstalling it, either because you experimented with it on a test box or on a virtual machine, or because the computer got hit by an earthquake, or simply because you didn’t know better, you may notice that the server continues to exist in the network’s Active Directory and shows up in the Exchange Management Console of your new exchange server.

This Post will help you clean those aftermath in your domain controller without having to reinstall your entire Infrastructure.

Removing AD Configuration

The list of Exchange servers on the domain is stored in the Active Directory configuration. Following steps below to remove the obsolete server:

1. Log on to the Domain Controller.

2. From the Start menu select Run and enter adsiedit.msc to launch the ADSI Edit tool

3. In the tool, right-click ADSI Edit and select Connect To…

4. In the Connection Settings dialog under Select a well known Naming Context select Configuration and click OK

5. Navigate to CN=Configuration [domain] → CN=Services → CN=Microsoft Exchange → CN=[organization] → CN=Administrative Groups → CN=Servers

6. Right-click and Delete the container for the offending Exchange server.

Removing Mailboxes

The Active Directory also stores information about the mailboxes that were located on the non-existent server. To remove these, perform the following steps in the ADSI Edit tool:

1. Navigate to CN=Configuration [domain] → CN=Services → CN=Microsoft Exchange → CN=[organization] → CN=Administrative Groups → CN=Databases

2. Identify the Mailbox Database containers that belonged to the lost server and Delete them.

Make sure that you are not deleting working mailboxes from the list, if you delete any by mistake you will have to delete & recreate the entire user using Active Directory Users & Computers. If there are multiple Exchange servers on the network, the Mailbox Database folders to delete can easily be identified by clicking on the folder and checking the Name field of the first entry, which should contain the name of the lost computer. Do not delete references to databases belonging to other Exchange servers!

Removing Security Settings

Now that the server is removed, some other settings need to be cleaned up as well. Perform the following additional steps on the domain controller:

1. Launch Server Manager

2. Navigate to Roles → Active Directory Domain Services → Active Directory Users and Computers [domain] → [domain] → Microsoft Exchange Security Groups

3. Remove the computer from the members list of Exchange Servers

4. Remove the computer from the members list of Exchange Trusted Subsystem.

If the computer was removed from the network altogether, you may also want to remove it from the Active Directory’s list of computers, as well as the DNS Server’s Forward Lookup Zones.

Please note that the above steps will work for Exchange 2007 / 2010 & 2013

2015/12/07 11:09 2015/12/07 11:09
Filed under Study


업그레이드가 되어서 플러그인이나 액티브X가 안되어 전사적으로 문제가 될수 있기에 정책으로 내려버리는 방법도 있습니다.

2015/07/27 11:02 2015/07/27 11:02