Main Area and Open Discussion > General Software Discussion

NLB and Load Balancing

(1/2) > >>

wraith808:
Has anyone had any experience with NLB?  Just wanted an overview from someone that I trust about what my NOC is getting me into.

(And the presentation is really boring and is an overwhelming amount of information to be handled in an hour presentation)

40hz:
It's a well established technique and it works quite well as long as it's set up properly and tuned correctly.

I'm more familiar (from a hands-on 'real' as opposed to 'lab' perspective) with automatic failover provisioning, which is a sorta distant cousin to NLB. The few clients I have that are into serious digital on-demand media distribution (a case where NLB really comes into its own) farm it out to distribution services and let them worry about it.

Sorry I can't be more helpful. SJ is probably the man to ask. Or Gothi[c].

Stoic Joker:
There's not a lot of call for that class of infrastructure in the SMB market around here. While I do remember quite a bit about it from the Win2k exam track (it was covered extensively), I don't really have any practical hands on with it.

However on the bright side, if you give us a thumbnail sketch of what is being proposed...40 and I can walk you through some of the theoretical stuff.

wraith808:
Basically, we have internal services that are being accessed from client machines, whether they be true clients, or loaders of information from external sources.  In order to increase our ability to handle the throughput from the clients, they are going to farm servers.

These services fall into four categories:
- WCF WF4 services
- DCOM components (that might or might not be enlisted in DTC transactions)
- Appian process endpoints
- ASP/ASP.NET services

From what little I understand, there's no central pooling mechanism- it's sort of peer other than the one specialized switch?  As I'm working from the development side, I'm most concerned about troubleshooting (especially in the case it is an intermittent problem which could indicate a problem with one of the nodes) and deployment and testing.

I feel a little better based on the last 5 minutes of the presentation, even though he didn't go into enough detail on that part, which had to do with gracefully removing a server from the pool and direct connection capabilities to a single server.  But in all reality, it was very overwhelming, especially as this isn't something that I'm going to be directly involved with on a day-to-day basis to bridge the receiving of the information to practical use.

40hz:
Yoikes! That sounds more than a little outside my area of expertise or experience. I doubt I could do more than toss out a few generalities about load balancing on that one as given. ;D

Stoic? You maybe wanna tackle it? :)

Navigation

[0] Message Index

[#] Next page