@JaredBusch said:

While my St. Louis region voice pulse numbers all show McLeodUSA Telecommunication Services, Inc

Of note with this. Only the voice pulse lines have intermittent issues calling some numbers.

Been trying to find a problem. This may be the cause.

McLeod is a third string voice provider in the grand scheme of things, but they should have plenty of interconnects to the POTS network to route that call properly. Almost everyone in VoIP buys numbers from someone else. That's why Level3 showed up for your other lines. It keeps the VoIP provider in numbers across the country and keeps Level3 in easy money.

For your problem, gotta narrow it down. IntraLATA, InterLATA? Specific exchanges or specific numbers? SIP to SIP? SIP to POTS? Cross carriers? Or the ever so fun forward to external number that happens to be international.

Ahh, right into my fun.

Short of having access to Telcordia, nothing out on the interwebs is 100%. Things get goofy when it comes to cell and VoIP providers as well, because lots of folks resell to Level3 which makes it harder to understand who is exactly supplying it and porting makes things crazy. Most of the databases out there are for the original providers. Most of it will be with ILECs and such.

This site appears to have things pretty close. It picked up my original SWB phone number and was able to narrow it down to the right VoIP provider and was able to identify my ported Google Voice number.

http://www.freecarrierlookup.com/

@Dashrender said:

Today people want access to their full bandwidth the entire time they are home, i.e. Netflix, etc.

Want in one hand, and shit in the other. See which one fills first.

This was the problem with the iPhone when it was released. Stupid, stupid people using the connection 24x7x365. This problem is much more apparent in wireless because you literally run into the laws of physics. Short of another revolutionary multiplexing process, e.g. from AMPS to GSM/CDMA, there is no way to pack more people into a frequency.

Pipes using physical medium run into the same problem. Even if they literally gave the entire DOCSIS bandwidth available to you, the upstream is constrained by the laws of physics in that you can't multiplex the connection on a DS3 at the headend any more than it is now.

I used to use an ISP that charged by the bit. People bitched and moaned about it because they thought bandwidth was free. Yeah, it cost a tiny bit more to use them, but I had one hop to the InterNAP backbone. 10ms lag to games, always available bandwidth, always good. This is where ISPs are gonna have to go to limit dumbasses leaving Netflix on all day long and not watching it.

@JaredBusch said:
.

People have always complained about the Netflix peering issue. That was a peering agreement dispute and not directly related to any kind of consumer network neutrality.

My "favorite" thing I've been seeing is people who don't understand peering and how it is monetized. Idiots have actually tried to argue that bandwidth is free, it costs just to get the pipe in.

Netflix et. al. want to be on-net with various ISPs because it costs them less than using the CDNs. ISPs don't want to do that because it would cost them more than just straight peering. But then they find that idiots leave Netflix streaming 24x7x365 which slurps their entire peer with Level3, costing them lots of money, so they have to weigh using them versus bringing in their direct peer pipe.

There is far too many out there who think that ISPs are looking to drive folks to their services by degrading others. Unless they offer a service superior to others, even with a degraded speed, people won't use it. This is why Netflix and Pandora are becoming ubiquitous because they offer a superior service versus others. If/when they were going to be throttled by ISPs, people would STILL use them.

The real winners of "net neutrality" are the level 1 backbone providers, Level3, InterNAP, hell even Cogent. They keep things the way they are now. Which isn't a bad thing. Although ISPs should have the ability to throttle things like that, because morons will still stream things 24x7x365 causing massive unbalanced traffic shaping. I guess coming up with a better routing protocol might be in order soon.

@scottalanmiller said:

@Dashrender said:

Wow... This could be a blow to the industry at large, we could see a fall in wages with this many people now in the market, depending on the positions.

Yes, although we don't know how many of those are IT people. IBM isn't an IT company, remember, they are a manufacturer. Mostly those will be manufacturing people. Only a very small number of IBM are in IT. When I was there we had about one IT person for every 500 manufacturing people (engineers, secretaries, factory workers, chemists, etc.)

WTF?

Global Services is probably the largest group of folks in IBM, they are pretty much all IT consultants. Rochester is almost all support folks for iSeries. Manufacturing was limited anyways in recent years, but most of their x86 stuff was built by contractors and such, like the plant down in Mexico that used to do x86 stuff. Semiconductor production was spun off to GlobalFoundry, continuing their dominance in the market almost everyone save Intel and Samsung have decided to exit. Even when they were in the x86 desktop market they let Lenovo do most of the work.

Don't know where you get your info, but it's way off.

@scottalanmiller said:

With FakeRAID not only can you, but often you will by accident. If the driver crashes, for example, you would suddenly have access to two (or more) raw drives with the system having no idea that there was supposed to be RAID there. With hardware RAID, that can't happen.

Love to see how that can happen, considering Intel/AMD/nVidia presents a single drive. So when you pop that drive off the machine, you are saying that you can access the raw file system of either drive?

Bullshit.

@scottalanmiller said:

@PSX_Defector said:

Unless it has a dedicated CPU, where do you think those cycles are coming from?

nVidia doesn't make dedicated RAID chips.

No Harware RAID maker does. They buy them from other vendors. Super high end devices like NetApp use custom ASICs so you could argue that they are chip makers. But normal hardware RAID uses ARM, Power or possibly Sparc chips. Nvidia, to do hardware RAID like AMD, would put the RAID processing into the RAID chipset.

Now you are confusing everything.

By your definition, southbridge/HT based RAID setup is ALWAYS "fakeRAID". A dedicated processor, be it from LSI, Intel, or any number of vendors, is ALWAYS hardware RAID. Anything configured at the OS level is ALWAYS software RAID.

Again, this is the same stupid arguments that have been going on since 2002. Calling it "fakeRAID" is just stirring up the pot.

@scottalanmiller said:

So, using SCSI or ATA commands, whatever is handy... how are you piercing the veil of hardware array encapsulation? Either it is a SmartArray P400 or a EMC VNX or an HUS whatever... how are you getting to the underlying devices past the array encapsulation?

You query it through whatever protocol you are using.

http://www.tinkertry.com/msminstallonwindows/

In dedicated hardware RAID, you talk to the card, the card tells you what you need to know. In chipset based RAID, it talks to the southbridge (Intel) or HT chip (AMD). Software RAID presents to the OS individual drives, it's ganged up logically to the OS specs.

@scottalanmiller said:

There is chipset RAID that does not use the CPU and is hardware RAID. It encapsulates. This is fact. The nVidia stuff that I have seen first hand uses the CPU (no hardware RAID) and does not abstract anything (no encapsulation.)

Unless it has a dedicated CPU, where do you think those cycles are coming from?

nVidia doesn't make dedicated RAID chips.

@scottalanmiller said:

@PSX_Defector said:

SAS is just SCSI commands over SATA.

Huh? There is no ATA protocol in SAS. SAS is pure SCSI.

SAS is an encapsulation of SCSI protocols over a SATA interface. Nothing I said is saying it uses ATA, although it can speak at that level.

@scottalanmiller said:

Like I keep explaining over and over, FakeRAID is really RAID. It just isn't hardware RAID. Chipset RAID is hardware RAID and not part of the discussion.

But everything you are talking about IS chipset RAID.

It's implemented in the firmware, abstracted by protocol, and uses cycles from the CPU on the board to perform calculations. It's hardware in the fact that it's controlled at the lowest level with regards to the hardware. It's software in that it is controlled within Windows through applications.

This is the same discussion everyone was having back when Intel first released it on their ICH6R southbridge.

It only seems as though Linux zealots are calling it fake.

@scottalanmiller said:

@PSX_Defector said:

So your "definition" of fakeRAID depends on if the system can pierce the veil of the abstraction? I can do that on LSI cards, Adaptec cards, just about anything.

Through SAS? No you can't. Show me that being done.

SAS is just SCSI commands over SATA.

Any and all cards using SAS can pierce the veil of disk abstraction to the disk level.

@scottalanmiller said:

Exactly. I know that AMD and Asus were making hardware chipset RAID (NTG buys them a lot via HP commercial gear) and that nVidia and Intel do a lot of FakeRAID trying to trick people into thinking that they are chipset RAID. And, as you can see with PSX, it is working.

No, because I've heard this same shit over and over and over again since the beginning of time with regards to "chipset" RAID. It's stupid back in 2002, it's stupid now.

So your "definition" of fakeRAID depends on if the system can pierce the veil of the abstraction? I can do that on LSI cards, Adaptec cards, just about anything.

Does it do RAID functions? Yes. Just because I can do some low level things like SMART status doesn't make it fake.

@scottalanmiller said:

Chipset RAID is RAID. Fake RAID is not chipset RAID. It is software RAID pretending to be hardware RAID. Hence the term Fake. This is a standard industry term, it is not a grey area or in question.

There are rare chipsets that do hardware RAID, like AMD. But they are very rare. Does nVidia make any of those?

Yes, depending on the specific southbridge, RAID is just another option for the chipset.

What you are ranting about is chipset RAID. Every RAID controller requires a "driver" in Windows. Does an LSI controller become "fake" because you load a driver to read it in Windows?

The abstraction is standard to all lower level RAID chipsets, be it LSI or Intel or nVidia.

Geez, not this shit again.

Chipset RAID is NOT fake. It does exactly what it says it does, mirrors or stripes datablocks between drives.

It's not hard to read up on these things.

Other than the 2003->2007->2010 method, the only thing I can suggest is migrating OUT of public folders into Sharepoint.

http://www.quest.com/migration-suite-for-sharepoint/

Although I haven't tried it in a while, maybe the Quest tools would be better for this situation anyways. 2003 to 2010 is a large infrastructure change, lots of AD prep and massive backend changes. Assuming since they are still using 2003 they are probably on a 2003 schema, so maybe using a better tool would be better for them.

http://software.dell.com/products/migration-manager-for-exchange/

Your thumbnails are way too big. I'm browsing it on a 1920x1080 monitor and they are huge, think what would happen with a mobile view. This is what is giving you all this scrolling. Tighten up Takahasi.

Your individual pages are way too compressed to one side. Again, I am on 1920x1080, but it seems you designed the site around 1024x768 hard. Make it dynamically render and adjust to resolution. Combine this with the thumbnail reduction, should be an easier method of browsing the site.

Your "free shipping" is where your logo should be, your logo is where your header should be. Free shipping notification should be relegated to the individual item pages and not front and center on the site. Everyone offers free shipping, it's pretty much expected nowadays. Although if you undercut others deeply enough, I would say it's worth a shot to not offer it. You only have 40 items or so, no point in trying to play with the big boys when you are just starting out.

That's my take on it.

@IRJ said:

I had no idea. Networking isnt my area of expertise. I am a Microsoft Server guy

I was just saying this to my "peers" in a certain subcontinent. insert dripping sarcasm

Knowing the fundamentals of networking allows all of us to be better engineers. Use the OSI model Luke!

Nothing stopping you from understanding bind or dhcpd. Grab a pfSense box and go nuts. It's when you start getting into the fun of deeper networking protocols, like NAT tables and BGP routing, is when you need to be much more dedicated to that discipline. Getting the fundamentals of bind down lets you really know how to troubleshoot a DNS issue. I'm called upon at work for being the "DNS guy". Even moreso than our DNS admins. dhcpd does lots of things and you can learn plenty by using it, even with regards to Windows DHCP services.

Never be afraid to grab something and f[moderated] around with it. Hell, I picked up a Mikrotik router a month ago, haven't touched the OS in a decade or more. After a bit of re-education, I think I can hold my own with it. Although it does load balancing very different than others do, much more traditional.

@IRJ said:

@PSX_Defector said:

@IRJ said:

I believe both DHCP and DNS are much easier to scale and mange on Windows.

Bwhahahahaha.

Say hello to bind and dhcpd. Bind will scale DNS faster and respond quicker than anything Windows can throw out. I've seen dhcpd hand out class A blocks and exhaust them and ask for more.

The tight integration into AD makes using it on Windows a convenience. Using it on anything else is always available.

I am not familiar with either, but the OP was talking about using DNS and DHCP on his router. Neither Bind or DHCPD.

What do you think every firmware uses?

Busybox, FreeBSD, WRT, even Cisco and Juniper use custom versions of this on their boxes. The class A DHCP server I've seen? On a Juniper.

Always be familiar with bind and dhcpd, never know when it might be necessary to fix it.

@IRJ said:

I believe both DHCP and DNS are much easier to scale and mange on Windows.

Bwhahahahaha.

Say hello to bind and dhcpd. Bind will scale DNS faster and respond quicker than anything Windows can throw out. I've seen dhcpd hand out class A blocks and exhaust them and ask for more.

The tight integration into AD makes using it on Windows a convenience. Using it on anything else is always available.