Thread started: Aug 21 2008, 5:29 PM EDT
Watch
Can someone please illustrate where Cisco's Solutions start and stop as related to Green IT? I am not talking hardware [i.e. the Nexus/infiniband/10-40g/VMWare/Consolidated Switching/Energy Efficient Ethernet/LB power down redundant firewalls and hand off to EMC/NetApp storage, and that it can be housed in an APC HD rack/cooling.] I am not asking about Cisco DataCenter energy audits/planning/education. I am asking beyond the hardware or consulting, and static efficiency identification tools, what other Green IT products does Cisco offer? Do you have any real-time datacenter energy management software suites? Are they holistic/agnostic that have visibility into facilities cooling and down to rack and x86 server level? I recognize the product hardware readiness and advancements but do tools exist to leverage them? For example the Tools to help change business behavior and processes like automatically shedding heat/power load on the nights and weekends and restart again when work load increases?
1
out of
1 found this valuable.
Do you find this valuable?
Do you?
Show Last Reply
|
|
Last Reply:
RE: Putting Teeth in DataCenter Energy Management (Question for Cisco)
By: ,
Aug 22 2008, 5:20 PM EDT
"jimmy, Thanks!
To really execute DC nights and weekend cost savings today as it was discussed in the TechTV presentation, it appears that a patchwork of scripts, software, tools, and careful QC testing would ensue. It may be possible to shed overly redundant DC loads, in the presentation EMC/VMWare said 40% of power costs can be saved by shedding loads during nights and weekends.
This could involve using VMWARE Virtual Center w/ Distributed Resource Scheduler (DRS), IBM Director w/ Power Executive, HP SIM w/ DSC/HP Insight Power Manager and Cisco's IOS Embedded Event Manager (EEM), to shut down load balancers, firewalls, routers, switches, servers, storage, vmware sessions, cooling, monitoring, as needed to achieve the desired goals.
Using VFrame power strip per outlet hard power shut-off seems to be a method of last resort and not suited for benign nightly use due to harmfull impacts of dirty shutdown while gear is doing something. A state full and graceful power down method is needed. Some of the overwhelming factors are synchronizing all the above tools for shutdown plus the NOC (Application, Server, and Network) staff will have a dozen SLA monitoring tools (ping, html gets, heartbeat) that need to stop monitoring the same powered off systems every few seconds. All those NOC alarms and SLA tracking/reporting tools will also have to be turned off and on once they come back into production.
A follow up question does Ciscoworks/View/RME tools let you schedule and manipulate power in mass to a group of devices all at once for nights and weekends?
" Very good call. Network Management has a ways to go to catch up with hardware. But isn't that always the case? Ciscoworks does not do that now. They have focused the most recent development stuff on Unified Communications. But I believe it will catch up and get there as time moves forward. So until then, I believe we have to blend elements to come up with a solution that fits. Personally, as I do like the power I get with EEM and VFrame. I have some perl scripts that I also have wrote for real time stats snaggin. Jimmy Ray
3
out of
3 found this valuable.
Do you find this valuable?
Do you?
|
