Questions regarding Veritas Cluster server
I have several questions regarding Veritas Cluster with replication. It shoud be only Windows Disaster/Recovery environment with two servers in 5 km distance with fibre optics between.
1. Is it possible to use only internal disk and not external storage ? Is it necessary to use external storage because of quorum shared disk, like it is in Windws 2003 cluster ?
2. Is it possible to use only Oracle database replication (Oracle DataGuard) and not volume replication ?
3. If I want volume replication, is Windows Softek replicator supported ?
4. How many network interfaces are needed on each server (and what bandthwidth) ? Is 3 minimum ? (1. heartbeat 2. replication and connection to each other 3. connection out for clients)
5. Is 1MBit or 10MBit bandthwidth enough for heartbeat network interface ?
Thank you for any suggestions
Juro
Comments
I'll take a stab at each question...
1. Yes, but you must use some form of replication, such as Volume Replicator avaialble with Storage Foundation. VCS does not rely on a quorum disk.
2. Yes, this is definitely available on the Unix platform but I'm not sure if it's available yet on the Windows platform. Check with your Symantec sales team for availability.
3. Softek is not currently supported with VCS.
4. VCS requires a minimum of TWO connections for cluster communications. A separate connection for replication is sometimes advisable, but not required.
5. Yes...LLT isn't bandwidth-intensive.
and is 1MBit enough for heartbeat ?
Thank you very much so far. I have one additional question - do you know if 1MBit is enough for heartbeat or is 10 MBit better ?
Anyway IBM is suggesting Softek replicator for UNIX with Veritas here: ftp://ftp.software.ibm.com/common/ssi/sa/wh/n/zsw0..., but anyway I needed Softek for Windows.
Thank you once again.
VSF HA 5.1 + VVR on W2K3 kovovyroba praca s kovom mreze brany
Where do you get 1Mb/s Ethernet?
I never even see 10Mb/s any more, let alone 1Mb/s. Or is this for channel allocation on a long-distance fiber link?
Yes, this is channel
Yes, this is channel allocation on long-distance. We have to pay for higher bandthwidth VLANs. So if it is not necessary, I would do it with 1MB only. I found here http://seer.entsupport.symantec.com/docs/288772.htm that 10MBit will be enough, but is it possible to do it with less :) ?
eh, Okay I will use 10Mbit.
VSF HA 5.1 + VVR on W2K3 kovovyroba praca s kovom mreze brany
I have never seen with 1MB,
I have never seen with 1MB, so would recommend with 10MB only....
Attempting for 1MB would be a try...
Gaurav
PS: If you are happy with the answer provided, please mark the post as solution. You can do so by clicking link "Mark as Solution" below the answer provided.
Multi-node failover
Hi,
Looking for references and the know how to configure multi-node failover scenario in VCS in an N+1 cluster.
Of particular interest is how I can set priorites per application or service group and how the cluster handles dynamic failover so that I don't overload a node in the cluster that has a failed over service group running in it under failover conditions.
Tks.
Hello, "Service Group
Hello,
"Service Group Workload Management" should be of intrest to you.. Please refer to VCS Users guide for this section..
You can find Users guide here:
http://sfdoccentral.symantec.com/Storage_Foundatio...
Hope this helps..
Gaurav
PS: If you are happy with the answer provided, please mark the post as solution. You can do so by clicking link "Mark as Solution" below the answer provided.
Solution
hello you can find the solution for the symantec tech support for this.
Would you like to reply?
Login or Register to post your comment.