• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Network pics thread

dashpuppy said:
DSCN0236.JPG
Click to expand...

That thing is begging to have caster wheels mounted on the bottom.

I put them on my small rack, makes it very easy to move around and make changes.
But knowing you there is probably very little chance you will be making any changes to that rack. :rolleyes:
 
Last edited:
MinPins said:
That thing is begging to have caster wheels mounted on the bottom.

I put them on makes it very easy to move around and make changes.
But knowing you there is probably very little chance you will be making any changes to that rack. :rolleyes:
Click to expand...

Ahem :)

All done for now,

DSCN0238.JPG


DSCN0239.JPG



This is for you Minpin's :)

DSCN0240.JPG
 
Dashpuppy, that's BEAutiful! Very very nice, super clean and just all around good work sir. Keep it up :)
 
dashpuppy said:
yeah it's part of my untangle project i'm playing with right now, ITS driving me nuts so i need to order new fans.

it was given to me, it's only a 24 port 10/100
Click to expand...

I have an 1810 on my desk(at work :( ), and besides its rather limited management......its fanless and 24 ports of gigabit goodness. its *ONLY* about 250 :(
 
As promised, a pic of my ASA. Grabbed an ASA5505-UL-K9 of eBay for 300 bucks. Brand new from the looks of it. Got it all up to the latest firmware, moved over my config from the Mikrotik on top to the ASA. Had to use ASDM because I was lost in the CLI when it got past the natting stuff. Moved over the SSL certificate as well. Also upgraded the RAM to 1 GB. May also upgrade the CF card, we'll see. So far I'l liking it a lot. I loved using Cisco's SSL VPN at work and I'm going to enjoy it at home now too. Beats having to setup an SSTP connection in Windows every time I wanted to VPN into my Mikrotik. I had to temporarily route my WAN stuff over my patch panel to some drops in my room. I wanted to get it all working rock solid before I moved it back downstairs into the rack.

swiftcurrent2012031200060.jpg
 
if anyone is looking for new cisco gear, I had good luck purchasing from lanstreet.com

they destoryed most of the pricing I could get on 2911's and the t1 cards I needed
 
sam-sgc said:
Details on the 5548 problems ??
Click to expand...

-Stacking, I think the stacking ring managed to get about 100KB/s tops..
-Vlans, sometimes they wouldn't stick to a port, 802.1q trunks were not stable
-Port channels, did not hold vlans ^
-Spanning tree, took out the entire switch for 15 seconds on average
-Ports also did not always turn back on after the device that was connected had it's own interface turned on and off. This caused some serious issues with my juniper and kemp clusters. Also caused problems with the HSRP that is on the uplinks

That's about it :p
 
Nacho said:
-Stacking, I think the stacking ring managed to get about 100KB/s tops..
-Vlans, sometimes they wouldn't stick to a port, 802.1q trunks were not stable
-Port channels, did not hold vlans ^
-Spanning tree, took out the entire switch for 15 seconds on average
-Ports also did not always turn back on after the device that was connected had it's own interface turned on and off. This caused some serious issues with my juniper and kemp clusters. Also caused problems with the HSRP that is on the uplinks

That's about it :p
Click to expand...

exact problem I have with the 2724's
 
Just to chip in on Dell Switches.

We have over 20 in operation in our facilities, mostly 35XXP, 5548P, 6224P and we almost never had any troubles with them. I've had some PoE ports fail on the long run, and some glitches that got corrected in firmware updates, but nothing major and recurrent.

Our core switch is a 6224P, doing L3 stuff (inter-VLAN routing mostly, other traffic goes to firewall, no dynamic routing) and it's been working flawlessly for the past year on our network of around 300 devices. We have another one in backup just in case it goes wrong. For the price they are, it's really cheap.

However, I know it's not carrier-grade stuff, but it's fine for SMB's.
 
network.jpg


Some of what the network engineer has let me borrow to help me study towards CCNA. He is still trying to find his 2960's from his move that will be added.

Having some trouble figuring out where to start :p
 
Console in, start locking down the router, configure your enable secrets, secure your ssh and telnet and console lines, give it a host name, setup NAT, maybe setup RIP between the two for starters. Go from there. As a visual learner I'm finding CBTNuggets, PacketTracer, and physical hardware way more helpful than any book.
 
Yea, started with that last night.

Odd how the books I found and have been using didn't teach me crap about breaking into the routers so I could factory reset them and blow away the old configs hah.

Was a fun time!
 
dashpuppy said:
Looking good guys,

Since we are experiencing 120k winds, i'm home for the day doing some training and in middle of that starting and working on this today too :)

Moving the rack from the rght side of my desk to the LEFT side, to do this, i had to pull all the cables under the des going from right to left and route them the opposite way.. Not to hard to do..

DSCN0236.JPG
Click to expand...

I would have mounted the switch backwards on the back rails to make it cleaner and easier to connect to the servers inside.
 
Got to unpack one of our new VNX 5300's today.

4 fast cache, 10 flash, 25 600gb, 15 1tb.

Projected IOPS - 30,700

 
Krazypoloc said:
I would have mounted the switch backwards on the back rails to make it cleaner and easier to connect to the servers inside.
Click to expand...

naw, don't want to.


Besides my servers and racked equipment is connected at the patch panel, i have a cable going from the patch to the back of the server..
 
I mean that SSDs are limited for the number of writes the can handle -- that's just how flash memory works. Lots of techniques exist to work around that limitation, but all of them introduce some amount of overhead, unpredictability, and bugs. In database applications that are write-intensive, these factors can cause issues in latency, performance, and stability, and can require careful tuning and planning to circumvent. Part of that planning involves evaluating vendors to see who best minimizes those factors, particularly when exposed to your application's access patterns.
 
Vito_Corleone said:
Forgive my storage newbness, but you say 4 fast cache and 10 flash. I thought flash drives ARE what's used for "FAST Cache"?
Click to expand...

15 drives total.

4 are in a RAID5 for FASH cache (64 KB chunks of data move to this during normal operation)

10 drives are in 2 RAID5 groups and act as Tier 1 Storage Pool (1 GB chucks of data more to this during normal operation)

1 hot spare shared within the 3 groups

Tier 2 Storage is the 600gb 15k drives, Tier 3 Storage is the 1TB 7.2k drives.
 
I mean that SSDs are limited for the number of writes the can handle -- that's just how flash memory works. Lots of techniques exist to work around that limitation, but all of them introduce some amount of overhead, unpredictability, and bugs. In database applications that are write-intensive, these factors can cause issues in latency, performance, and stability, and can require careful tuning and planning to circumvent. Part of that planning involves evaluating vendors to see who best minimizes those factors, particularly when exposed to your application's access patterns.
Click to expand...

With the latest wear leveling techniques, overprovisioning, error correction, and write applification, SSD write endurance is much better than it used to be. While it's true there are still limitations, it's not as prevelant as it used to be and if you look how EMC and other vendors leverage their flash, you can see it's usually not for consistant write usage.

The useage of flash is granular and spread accross drives. Most likely you are utlizing your flash in chunks and in EMC's environment 64KB for FAST Cache and 1GB for FAST Tiering. The whole idea behind this is to utilize flash only when needed, ie bursty I/O or "hot" data.

Some environments recommend an all flash pool or raid group like Replicas for View Linked Clones for VDI, but that's all READ based.
 
Since I can't take a photo of my network... here's my sample photo. The environment mostly consists of Dell R5400s and numerous EqualLogic 6500s. It's all managed by 4 2960Gs, no need for a router since there's no outside connectivity.
 
ben chi(f4) said:
Since I can't take a photo of my network... here's my sample photo. The environment mostly consists of Dell R5400s and numerous EqualLogic 6500s. It's all managed by 4 2960Gs, no need for a router since there's no outside connectivity.
Click to expand...

Whats the tipe of service that so many servers provide without outside conectivity? :confused:
 
Lots of Government stuff is like that. They even host IE on a citrix farm outside of the building / network for internet connectivity!
 
Nocturn3Ev1l said:
Whats the tipe of service that so many servers provide without outside conectivity? :confused:
Click to expand...

Goverment as Jay mentioned. At my job, we are building a $800 million dollar comms job and on the 2nd network, its completely isolated from the rest of the internet.
 
Vader said:
With the latest wear leveling techniques, overprovisioning, error correction, and write applification, SSD write endurance is much better than it used to be. While it's true there are still limitations, it's not as prevelant as it used to be and if you look how EMC and other vendors leverage their flash, you can see it's usually not for consistant write usage.
Click to expand...
Yep. In this case, you're paying a high premium to get around those limitations. If that works, then great -- but it's funny that you're getting neither the full benefit of SSDs nor the stability or predictability of spinning drives. The complexity is pretty immense.
 
You must log in or register to reply here.
Back
Top