I believe you read it correctly then... I must establish myself as the Dominant AlphaGeek when it comes to system engineering...[/quote]Dashrynn wrote: all i read was "im more nerdier than the rest of you and heres a long list to prove me right"
I can agree to this. Great equipment, not the most profitable business.lamont wrote:NetApps are awesome, but 'spensive.
One guy on IRC commented one time that EMC shops are full of people who are broke and are constantly responding to pagers, shops with NetApps are equally as broke, but they get to sleep all night long.
They're the only piece of equipment I think I've come across that I think is really four 9's, maybe five...
How dare you...I'd say a Netscaler 17000 instead of BigIP.
this I STRONGLY agree with. The cost of running this system on EC2 with their new database application would be comparable to what they are paying now for this lousy service. I too believe I could build a compression schema that would allow better performance from my home server than what Dreamhost is giving these guys now.I don't know what kind of compression they're running, but at work we only get about 10:1 compression, tops, which is still only about 100Mbps peak utilization of the network to any virt host.
We're thinking of going to 4900M's or equivalent switches from Juniper and running a complete 10Gig access layer now (which i think is back in the regime of massive overkill).
Anyway, if this is a hosted service, then it really should be changed to a different hosting service... I could probably run this better out of my server at home on a DSL line based on how well its been doing lately, and i figured this was just Josh's or someone's pet project -- but if this is a pay-for service, that is kind of bad...
Actually, thing to do would probably be to just run it in EC2 and rent space from Amazon... That's also virtualized, but the servers there are not as oversubscribed and you really get a guarantee of how much RAM and CPU horsepower you get...
Yeah, but they do manage to make the big bucks off of it.... sigh.... 0]Yeah, clearly if you do get smacked by a worm, it isn't going to obey the kinds of courtesy in terms of hits per second that the Google webcrawlers at least attempt to...
And its Russian mafia and other criminal organizations operating outside of the reach of USA law enforcement which is responsible for a lot of the blackmailing operations and attacking people for profit...
You may have met your match my friend.... Well... I usually try not to spout off about what I know, so maybe not. And I'm more network engineering, and kernel level platform integration....I believe you read it correctly then... I must establish myself as the Dominant AlphaGeek when it comes to system engineering...Dashrynn wrote: all i read was "im more nerdier than the rest of you and heres a long list to prove me right"
i believe you and defied and airsix will dominate but you will have to dominate each other.[/quote]Dashrynn wrote:[quote="lamont]I believe you read it correctly then... I must establish myself as the Dominant AlphaGeek when it comes to system engineering...Dashrynn wrote: all i read was "im more nerdier than the rest of you and heres a long list to prove me right"
Silly billy, that website doesn't exist! 0]Norris wrote:They are copying and pasting off of
http://www.rantaboutnetworksserversandD ... eapair.com
:roll:
Code: Select all
]# whois www.rantaboutnetworksserversandDBslikeyouhaveap air.com
[Querying whois.internic.net]
[whois.internic.net]
Whois Server Version 2.0
Domain names in the .com and .net domains can now be registered
with many different competing registrars. Go to http://www.internic.net
for detailed information.
No match for domain "WWW.RANTABOUTNETWORKSSERVERSANDDBSLIKEYOUHAVEAPAIR.COM".
Actually I think that most storage like this is about to get severely disrupted. Google, Amazon and Yahoo all built their storage based on GoogleFS-like clusters of cheap commodity servers that push the $/GB down to about 2x the cost of picking up the drives themselves at fry's (4x the cost of spindles at Fry's if you duplicate the data). EMC and NetApp can't compete with that kind of cost structure.defied wrote:I can agree to this. Great equipment, not the most profitable business.lamont wrote:NetApps are awesome, but 'spensive.
One guy on IRC commented one time that EMC shops are full of people who are broke and are constantly responding to pagers, shops with NetApps are equally as broke, but they get to sleep all night long.
They're the only piece of equipment I think I've come across that I think is really four 9's, maybe five...
Ah, I didn't realize that's where you worked.How dare you...I'd say a Netscaler 17000 instead of BigIP.
The Netscalers are actually really good machines, and they are our biggest competitor (Juniper was.... was...).
We have them owned in the platform market, but if I recall, their pricing is better. They are, however, heading down the same business model path of many other shops, where they lowball as much as they can and end up losing the developers who want the big bucks, because they can no longer afford raises. But, like I said. They are great competition, and are very good at what they do.
Yeah, I just responded to the Ivar's thread and its still running slow, and its just this site (well, and scubaboard, but that is always slow too).this I STRONGLY agree with. The cost of running this system on EC2 with their new database application would be comparable to what they are paying now for this lousy service. I too believe I could build a compression schema that would allow better performance from my home server than what Dreamhost is giving these guys now.
On the networking front, once you get into BGP then I'm lost. I've got a reasonably strong RFC-level knowledge of OSPF, but don't have any actual keyboard-level knowledge of that. When it comes to edge-networking of servers, I become much more of an expert -- so, I debugged things like the way that full-NAT loadbalancers violate the PAWS/TCP Timestamp RFCs and what that can do to very high TPS servers as you start reusing TIME_WAIT sockets and the RSTs that can get sent back by servers via just spelunking in the kernel code looking for where the linux stack will send a RST in response to an initial SYN. I also did most of the server-level security at Amazon since the security team there was mostly SDEs and didn't like doing SA work (I had an open invite to join them, but never did, since most of their day-to-day crap was just phishing site takedowns -- which got outsourced after I left), and did most of the global server configuration management when I was there -- which went from 400 machines when I was hired, to doing config pushes to 30,000 servers when I left. So, I've updated something like 15,000 servers running RH7.2 and old ssh 1.2.27 sshd binaries to openssh-4.0p1, along with building most of the gold standard image of what Amazon was using (patching the config management infrastructure they had when I got there to support the concept of a global image and then doing 80 out of about 100 of the commits to that global image when i left). In addition to that was just handling all the edge condition escalation for things that wouldn't happen once in 10 years at a site with 200 servers, but which happen once a month when you've got 30,000 servers.You may have met your match my friend.... Well... I usually try not to spout off about what I know, so maybe not. And I'm more network engineering, and kernel level platform integration....
