@pauska and typically newer 'features' come to the oldest type first. For example, we use a lot of wavelength-specific transcievers instead of 1550 for WDM, and initially we could only get wavelength-specific in Xenpak, but not X2
X2 and SFP+ are two different things, you can't plug one into the other
The TwinGig SFP Converter Module (Figure 2) converts a 10 Gigabit Ethernet X2 interface into two Gigabit Ethernet Small Form-Factor Pluggable (SFP) ports
Cisco SFP+ Transceiver Modules for Cisco OneX Converter Module
Don't expect to buy that X2->SFP converter and plug it into any random piece of hardware though
and a SFP+ is the same form factor as an SFP, just capable of 10G data rates (but obviously the hardware it's plugged into has to be 10G capable if you want to use it in that fashion - compared to an XFP, which is a little bit bigger, and obviously only goes into XFP-compatible cards
"The Cisco OneX Converter Module is supported on all 10 Gigabit Ethernet modules of the Cisco Catalyst 4900M. The Cisco OneX converts an X2 interface into one SFP+ interface and uses the full range of Cisco SFP+ transceivers"
@Chopper3 You're talking about the FlexFabric switch? If so, I think you're mistaken, only because when we plugged in 1GB sfp's in some of the flexfabric ports, it didn't work, and the HP guys said we needed the Flex10 instead. Or was it just because we were trying to do 1GB instead of 10GB?
@PeterGrace yes I was, ports 1-8 take 10GbE SFP+'s, ports 1-4 will also take 2/4Gb FC SFP's and 8Gb FC SFP+'s and ports 5-8 will also take 1GbE SFP's - is that ok?
yeah well, so nexus is out of the question for us anyways, 5k's are way too expensive.. so either cisco 4900m's, or hopefully HP launches something.. but it's not going to be before 2013
@pauska I'm still confused and I'm trying to make sure I answer the right question. 10GE is 10GE is 10GE, anything that talks it can be plugged via fiber into any other thing via fiber
@pauska controller cache is probably more efficient (localized to the stuff being accessed by one box rather than trying to handle the entire SAN), but a write-back cache inside the SAN can be helpful
I am running a web app on a LAMP server with 10GB RAM and quad cores. It serves about 170 login requests per second.
Is this number right?
Is there any way to inrecrease it?
@Zoredache I think it's a unetbootin issue or something about this particular stick (HP Product) -- Stubbornly insists that it's "Missing operating system" when you try to boot from the USB stick
if I can't beat it into submission I may harass the U&L folks
@Chopper3 It's weird, I swear Mark left a note on his profile telling him to get his flag weight up and a bit more meta activity, and it's gone from 140 > 500+ ...although there seems to be a lot of comment culling
@Basil or you can run stats which will offer quite a lot of disk and system level stuff and WAFL metrics you need to pull from SSH into a flat file and run gnuplot or something like that
We have data center running more than 2000 servers. We are planning to shutdown the servers on a weekend for cost cutting purpose. How can we calculate which servers are consuming more power in the data center so that we can shutdown those many servers only . What is the best way to determine pow...
I think they're getting their field circus monkeys from the OEM vendor's zoo
. . . I hate the inconsistency of Linux, I really do. I have done nothing different to this USB drive than the last 3 times I tried to create it. It works now.
I am left to conclude that in order to produce a bootable USB thumb drive one must simply try three times.
My build server is Caesar. The USB thumb drive is the crown. Later tonight I'm going to sneak back into the office wearing a toga and stab the build server repeatedly.
@WesleyDavid did you ever get that spiffy frontend running for your 10k challenge? I see you recently posted the results, but I didn't see a link to a website at all.
@pfo So back to storage: from what I can tell, the DS8 seems to not scale well. Both the USP and VMAX support larger configurations... is there something I'm missing?
I am running a web app on a LAMP server with 10GB RAM and quad cores. It serves about 170 login requests per second.
Is this number right?
Is there any way to inrecrease it?
