Search results

Been awhile, but this is still available.

Juniper SRX1500 firewall (4x 10G SFP+ interfaces), some scratches and non-standard components (read below) but otherwise complete and functional with dual PSUs and PIM slot blanks. I'll accept paypal (preferred) or venmo. Price includes tracked/insured UPS shipping, and I'm happy to take offers...

@eduncan911 : sorry, I can’t lend these out. Otherwise: combo still available, dropped price on the package to $1250 usd shipped.

super-interesting... to be clear (just in case), after flashing the stock firmware I no longer need to have the recovery mode jumper to boot. I had three systems show the same behavior with these cards - two H12 boards (H12SSL-nt and H12DSi-nt6) wouldn't get past pcie initialization, and an...

The first time I went through the process, the card stayed in recovery mode across reboots, even though I had taken the jumper off while it was still plugged in. I had to physically remove the card from the slot or remove all power from the system and let caps bleed to get it out of recovery...

These oracle ones work after performing a recovery mode (jumpered) flash-with-stock then power cycle: Mellanox CX556A ConnectX-5 EDR 100GbE PCIe NIC no PCIe Bracket 7290108480320 | eBay With the Oracle firmware the cards prevent the system from booting, unless the recovery mode jumper is in...

oh, I actually just managed to grab a couple oracle CX5's off ebay for not /too/ bad a price, so no longer interested in that side of things (will edit the main post momentarily). Apologies for missing this earlier. One combo still available, and added a ROMED8-2T that I also briefly used...

One combo sold, updated main post.

I set these prices based on ebay completed listings (for the CPUs), and below retail for the motherboard. Note that there is a price gap between the version of these motherboards with 10GbE and the ones without. But to others, please do make an offer if you’re interested in purchasing...

lowered prices

AMD EPYC Rome 7402P - $800 Tyan S8030GM2NE-2T (the dual 10GbE variety) - $650 Or take the motherboard with the 7402P installed for $1250 Decided to move to a dual socket board, so these need to go. All items were working fine when I pulled these out of the rack. I’m lazy and would love it if...

worthy of note: these boards do not support lrdimms - i found that one out the hard way.

Noticed the product page was updated to indicate Milan support:

I also just realized the version numbers are different for equivalent bios releases across the S8030GM2NE vs S8030GM2NE-2T (<- which I'm using). Apologies for any confusion around the version numbers in my posts!

Tried out the new V2 bios, and it resolved the "BAAD" post code issues described above. Both systems up and running. Note that it does reset bios settings as part of the update (dangerous because it set fans back to 30% PWM, decreasing the effectiveness of my watercooling setup), and in my case...

Noticed this in the 3.00 BMC firmware notes and got hopeful. Updated BMC on one of my boards then tried to update to bios v1.01, same brokenness as earlier, boot hangs at post code "BAAD" My response from Tyan support back then was "we cannot reproduce this in our lab". Flashing back to...

You’re right - 1.20 is the one I was thinking of, not 1.30. The images posted by fake-name are the correct ones. It ought to be possible to flash the nic bios without fully desoldering by using the clip-on adapter - I just haven’t ever had good luck doing so on other boards. lifting the vcc...

this one from amazon ought to work. What I've learned with TL866's is to try hard to get an official one, then flash and read a couple times to make sure it's writing/reading consistently. I've seen rumors of pirated ones, where updating them with the official bios will intentionally make them...

It’s worth trying an upgrade to the 1.30 bios posted earlier in the thread. Unsure if others figured out how to do it from esxi / linux, but a TL866/similar programmer works well. On one of mine I never did get the NIC to work on one of the nodes, so YMMV. I spent some vacation time today...

bonus pictures traces under the fpga (removed from one of the DOA boards, and not cleanly): broad area around the fpga and jtag header, with all of them (and the problematic mux) removed: glamour shot of my new rpi400 being used as a jtag debugging console: