Hi! I would like to suggest a 2nd kit for expansion. Let’s say you invest in the HL15, and all slots are full, and you love your HL15. It would be realllly cool if you could order a HL30DS (HomeLab 30 Bay Disk shelf) with the same external size as the HL15, but having 2 rows of disk slots, 2 backplane and 6 fans and a PSU, with something like Supermicro does to connect the PSU, power button, etc. The 2 backplanes would end up at the rear or that hypothetical case with the big disk cable for an external cable that you could put to connect to the main HL15 unit “brain”.
THIS! I love this idea of a disk shelf!
So a second chassis altogether as a JBOD? I’d be interested to see this, as again most of those out there that are available are enterprise, loud, power hungry, and sometimes proprietary.
Yes, that was my idea. A way to expand it.
With their choice of vendor neutral backplanes and SAS controllers already (LSI) this seems a very viable path! To be honest, one could add a PCIE card with a SAS expansion cable between the initial HL15 and a separate barebones HL15 chassis, by installing a small JBOD board in that second chassis.
Now granted, this would only add another 15 disks in the HL15’s current offering. However, it does show that your idea is already possible with a little extra work, and therefore may come to fruition if given enough demand!
Another upvote for this idea! I think even those homelabbers who aren’t considering the HL15 for various reasons, would find such a product attractive. And I suspect that the incremental development + material cost over the HL15 would not be that significant.
I was ready planning on something along these lines with a 16 bay chassis I have (after modding the dam fans), I recently then found art of servers jbod expander which should make it a lot easier.
Though this will not line up with 45 drives ethos of direct connected drives assuming that the 45 drives team release a HL30+ in the future it would not be hard to homebrew the DS
I wonder how easy it would be to mod the HL15 chassis to a “HL30DS”. It seems it would be pretty straightforward to DIY it if 45drives sells the HDD backplane as a part that one could order. Lets see what’s available on their e-store once it goes live.
this is basically my deciding factor. I’d prefer a 30 bay unit with the option of a 30 bay expansion but if i knew I could start filling a 15 bay server with 30 bay expansions on the horizon i would absolutely pull the trigger today
@anon You can reach out to Sales and ask what a bare Storinator Q30 would cost (if you were going Chassis Only, otherwise you can just use the shop)?
45Drives has to the best of my knowledge never made a JBOD “Expander” so I don’t think expansion shelves are on the cards, particularly because expansion shelves would not be a “Direct Wire Architecture” which has been a pretty pivotal part of Storinator’s approach since Day 1.
This is exactly what I was thinking when I made this post.
I agree with everyone else on this thread that this is both a phenomenal idea and sorely needed! And I think the design doesn’t necessarily conflict with the “direct-wired” approach. Right now the HL15 has a backplane that ultimately interfaces with the motherboard through an LSI SAS HBA. If you essentially treated the HL30DS as a 2nd backplane and used a LSI SAS HBA with external ports rather than internal ones (a la the LSI 9201-16e or similar), voila you essentially have maintained the same “direct wire” architecture.
I think this “HL30DS” type product is a much needed entry into the homelab category / product tier, and would provide a great entry point for the 45Drives brand to customers who have an existing storage solution that just needs more capacity (for the time being ;).
FWIW, they do have a video explaining their resistance to this type of architecture, at least for their enterprise clients;
I don’t agree with all of the points
For HL, I think it would be a good idea, rather than have to source a solution from used Supermicro, Netapp, Dell or other second hand enterprise disk shelves. I can see how they might not want to do it as it would conflict with the direction they are giving their enterprise customers (and make said product available to them). But, OTOH, it would not be an enterprise class solution with redundant PSUs, etc.
Environmental issues can creep up fast. Power/Cooling always needs carefully considered. For me, I’d rather build a second HL15 and take the path towards distributed.