Remote Desktop session disconnects or hangs..
Discovered an issue with Microsoft patch KB969947 installed on a machine with an NVIDIA Quadro FX570 video card. So far I have confirmed that with this combination you will no longer be able to remotely access your server or workstation via RDP unitl you uninstall this patch. I even had a machine completely lockup up connection.
multi-protocol san backend device. 4 HP DL260 g6 with emedded TOE and additional dual nics. to be used for SQL Cluster and NLB web/app servers.
replacing a mixture of in-place proxy servers and content filtering devices with Websense. So far so good. I really prefer the port monitoring / out-of-path deployment of Websense instead of the traditional in-line monitoring solutions. I will post up more when I get further ahead.
Scenario: two Silver-peak NX2600 devices placed on each end of a dedicated point-t-point T1. Major traffic type: Netapp Snapmirror, SQL replication & Citrix ICA. Devices are configured as an out-of-path deployment and we use map list to direct interesting traffic to the Silver peak devices for accelleration.
Results: fabulous results. I am seeing upwards of 85% reduction on all types of traffic. Snapmirror deltas are being completed in 8-10 secs every 10 minutes. SQL replication jobs normally with a duration of 2 hours are now being completed in 20 or so minutes. Very happy with the results. Price was very reasonable compared to other options.
A situation arose today that required me to manually remove metadata from Active Directory relating to a failed domain controller – dcpromo failed.
Sooner or later you’ll have to perform one of these. When you do just make sure that you’ve made every effort to fix any errors related to your failed dcpromo. Make this process your last resort.
The Ntdsutil.exe is included with Server 2003 Sp1 installs.
Log into the domain controller and from a command prompt type the following (commands are in bold)
1. ntdsutil
2. metadata cleanup
3. connections
4. connect to server servername
5. quit
6. select operation target
7. list domains
8. select domain number
9. list sites
10. select site number
11. list servers in site
12. select server number
13. quit
14. remove selected server
This will remove all metadata from Active Directory. As best practice you should remove the cname record in _msdcs in DNS. Also remove any DNS records that refer to the DC you are removing.
Also, go through Sites and Services and make sure there are no trace object of the DC you removed. If there is simply delete it.
When you have to increase the size of a source volume in a Snapmirror relationship you have to increase the destination volume as well. Here are the steps.
1. Break the mirror relationship (this makes the dest volume writeable)
2. Increase dest volume size to equal or greater than the source vol
3. Then execute the snapmirror resync command (this marks the vol as read-only, makes in a snapmirror dest vol and starts the transfer)
I chose NFS hands down. Here’s just a few reasons why.
- Easier to manage in a VMware/Netapp shop. Each VM folder containes their respective VMDK files.
- Volume space can be expand on the fly as desired and immediately available without a rescan or dealing with extents.
- Easy Snapshots recovery of VM’s. drag and drop.
- Because I have the opportunity. some shops dont’ have the opportunity because of budget. NFS is an added licensed feature.
Vmware / Netapp Disaster Recovery Using Snapmirror -
My setup:
Two datacenters connected over a T1.
Each datacenter run’s 3 esx host connected to Netapp FAS2050 via NFS.
Silver peak WAN Opt devices sitting at each end.
Vmotion/DR/Snapmirror fully implemented.
50+ vm’s running apps and sql servers.
ASIS nightly.
I organize all of my VM’s on each volume based on level of availbility. I configure my snapmirror to refect it. example: prod1_vm = 200 gb (snapmirrod every 10 minutes. prod2_vm = 200gb (snapmirrod every hour), prod3_vm = 400 gb (snapmirrod every four hours)
Questions or comments welcome.
Steps to build a new SPS front end with new Config_DB attached to your restored or existing content_db
NOTE: Always make sure you frequently backup your sharepoint content database.
1. Build or Power On a new Sharepoint web server.
2. Create a new Farm by creating a new config db (Sharepoint config wizard)
3. Create a new web application with a new content database ( Application management)
4. Remove newly created content database.
5. And then run this command to attach to your existing database.C:\Program Files\Common Files\Microsoft Shared\web server extensions\12\BIN>stsadm -o addcontentdb -url http://sharepoint.domain.com/ -databaseserver sqlservername -databasename wss_content
Your backup up!
1. The creation of file shares is dependent on Microsoft Windows Server Service. If the iSCSI drive is not present at the time this service wants to run you will have issue. To make this service dependend on the iSCSI service do the following:
On the host go to cmd and type: sc config LanManService depend= /MSiSCSI
2. In its current version, MS iSCSI supports only Basic drives, not dynamic. If drives are not being recreated after a reboot this is most likely the problem. Change your disk back to Basic.
3. Make your volume peristent. In the iSCSI initiator properties go to the “Bound Volumes/Devices” and bind all devices.
4. The Target must also have been added as a persistent target by selecting “Automatically restore this connection” in the Logon to Target dialog.
5. MS iSCSI Software initiator works great within Guest VMs. Just make sure you treat it like a physical host and dedicate a NIC for this traffic. So in ESX create a vSwitch for VM data storage and add a pNIC. In your VM add a vNIC and connect it to the vSwitch you just created.
