Alex Gerulaitis
Forum Replies Created
-
[topher rehpot] “Yes it is possible without san setup as this forum conversation states
https://reduser.net/forum/archive/index.php/t-38307.html“
Do you mean this post?
[Brandon Kraemer] “We have been able to stream M_proxies off of a Promise RAID, served by a X Serve in a NAS configuration via gigabit ethernet for cutting to multiple workstations. This doesn’t require FCServer or a SAN setup. We don’t rely on this, typically cut with DAS (direct attached) RAIDS, but it can be done.”
Afraid it’s a whole different animal vs. what is being discussed here. Any DAS or SAN can be “shared” using a NAS front end, but this entirely changes performance expectations.
The thread discusses whether a TB storage box can be shared among multiple TB-equipped Macs at full TB speeds without the added costs and complexities of using a NAS front end or SAN software.
— Alex Gerulaitis | Systems Engineer | DV411 – Los Angeles, CA
-
[Chris Murphy] “If either Q or P parity chunks are being corrupted somehow, then absolutely data chunks are being corrupted. “
Not sure I get it. If we’re not looking at a possibility of simultaneous and independent corruption of two data blocks within one stripe… If one parity chunk is corrupted, you could always re-create it from the other parity chunk, and the data?
Isn’t 6 equivalent to a 3-way RAID1 for the purpose of data recovery? One copy is corrupted, you’re not sure which one is good out of three – just check which ones match, assume those ones are good, discard the mismatching one?
Applying that to 6: calculate P and Q again from the data chunks, compare them to existing P and Q; whichever one mismatches – re-write it, and Bob’s your uncle?
If the data chunks were corrupted, then P and Q would still be healthy and you could re-create data from them, supposedly?
-
[Olly Lawer] “We did some tests based on a typical example of our work (knowing that AE performs better with higher RAM in some cases and cores in others).”
Can’t beat that. Best value is defined as unit of money per unit of productivity; if your tests show that you get more value out of 6-core configurations in your workflows, who could possibly argue with that? 🙂
-
10GbE for TB-equipped Macs: Small-Tree ThunderNET TN10-1P — 1 Port 10GbE Cat6 to Thunderbolt, $945
Switch:
– Netgear ProSafe M7100-24X 24-port 10GbE Switch, P/N: XSM7224-100NES (SRP $7,990, street around $5.5K)
– GS752TXS (or GS728TXS) – if you need stacking, scalability, more than 24 10GbE ports, a single management point, etc. Otherwise – either the M7100 above, or:
– NerGear XS708E 8-port 10GbE switch, around $1KCat6 is fine for wiring unless there’re unusual length / configuration requirements.
Am I hearing you right that your server is a Mac Mini with TB-to-SAS storage?
— Alex Gerulaitis | Systems Engineer | DV411 – Los Angeles, CA
-
Shane’s response is right on the money yet the specs don’t list 4K as a source for downconversion, which is why I think he said possibly. If it can – then it’s probably the least expensive for the purpose.
I don’t think it has anything to do with TB or DP per se – only with device’s hardware down-conversion capabilities.
— Alex Gerulaitis | Systems Engineer | DV411 – Los Angeles, CA
-
Small-Tree GraniteSTOR TitaniumZ-5 is $7.K for 10TB.
Other than lower cost NAS and SAN boxes from vendors specializing in video (GB Labs, EditShare, SNS, Maxx, etc.) you could look either at DIY or generic NAS from the likes of QNAP, Synology, Thecus.
-
[Kannan Raghavan] “Two options I found were the Editshare Field 2 or the Editshare Ultra. It is a bit expensive but both Avid MCs and FCPs can share media.”
Have you looked at SNS Evo by any chance? File/volume, Avid bin locking, compatibility with FCP(X) and Adobe CS/CC, some cool things like ShareBrowser (lightweight MAM)… Starts at $7K I believe…
— Alex Gerulaitis | Systems Engineer | DV411 – Los Angeles, CA
-
[Tim Jones] ” there’s definitely something in the USB 3 layer that’s causing a reduction in I/O:”
Indeed. Areca’s new TB/USB 3.0 4-bay box (5026) is showing impressive numbers via USB 3.0: 350MB/s in RAID5. It’s not cheap though – and I am not sure how it achieves these numbers, there’s still some black magic in USB 3.0 adapters and their configuration: e.g. UAS is supposed to boost performance quite a bit, but very few boxes and adapters support it.
— Alex Gerulaitis | Systems Engineer | DV411 – Los Angeles, CA
-
[Iljitsch van Beijnum] ” lowering the MTU makes no difference at all to the sending TCP stack, which is in charge of setting the transmit rate”
I meant lowering the block size (below MTU) before it gets sent to TCP. Eliminate TSO overhead.
-
Would artificially setting the packet size below MTU size help figure if it’s TSO that’s the problem? E.g. doing a disk or iperf benchmark with block size at 1K?
Just wild guessing that this might help limit or eliminate segmentation and if we see less choppy bandwidth (although perhaps much lower due to high latencies) – that will pinpoint probable cause?