Absolutely true!
To be honest, I am absolutelly with the 8GB on board, BUT:
I would like to have an option to change the emmc whenever bad block appear or needs of an bigger capacity - as I already have 256GB emmc chips and I am not able to use them
Installation of any OS on the emmc will be much easier - it is not really big challenge whenever you know what to do, butā¦
What about at least 2 more sfp+ net-ifs? What would it needs?
9k MTU - does it really do not needs any hw changes and only sw changes are required?
rss/lro is available already
more powerfull cpu - does there is any available/usable
R4Pro is same SoC so it has same count of pcie lanes but it is basicly switchable by a gpio between slots. But i have not tried it yet,still try to get networking working on mxl switch.
The question is why is a more powerful cpu neededā¦where is the bottleneck?
Btw. The schematic does not match the board i haveā¦both sfp/10g-phy combo ports are realized with an ethmux
I assume, that with MTU at 9k will be more efficient in the high performance tasks on sfp?
Does it is possible that 2k MTU causes avg load on all 4 cpu cores to 70% and therefore limit the performance of writing data through sfp port?
Does there will be still enough cpu power in cases like fe:
primary server functions nfs, samba, tftp, iSCSI for disc-less stations on SFP and even ETH ports
using RJ45 ports with both of the sfp ports in case of night backups, rsyncing or something like that?
managing of raid5/6/10 arrays, including re-shaping, re-checking on background, and also working in degraded mode?
Hello! Popping in here because I would like to know about a potential release dateā¦
Currently I am using an ONT from my ISP that plugs in via SFP but my DrayTek router is being a bit of an issue - and a bottleneck. So I am very much looking forward for this board.
Is there any availability ETA or something? This, with OpenWrt, seems like a perfect fit for what I need.
If the R4Pro isnāt really ready and should be avoided for now, what would be a suitable choice of mainboard instead? I purchased a Turris Omnia several years ago and it was a good unit at that time, just looking for something updated to serve the same purpose. Thanks!
i have a really early board, i guess most of the issues are known so far and are fixed when the board comes into saleā¦it is not really relevant at this time for end-usersā¦you only need to wait some hardware iterations and software development before getting such board except you also want to TEST and DEVELOPING (at linux driver level, not userspace) itā¦it is not for using in production environment yet
The pin-header is for gpio,you cannot short any othr output to it. That would require a led headerā¦that is no bad ideaā¦so instead of soldering leds,add a pin header for it where per default a led stripe is plugged which can be removed and replaced by cables with leds to route them to a case.
Iām not sure it mpcie and m.2 have generic āactivityā pinsā¦afaik sfps do not have them.
My goal was not discuss here the accessories for the board, but changes for the next version of the board which could be benefitial. I can see on-board leds usefull for those who are using original-like case for their projects.
For me are all on-board leds which I am not able to replace for the external ones nothing more than another problem to solve.
I need my own case which have to be much bigger than the original one (more than 30x30x30 cm), and now I have to de-solder all on-board leds and replace them by cable or connectors to be able to connect external leds in front of the case.
I believe that this details are important for everyone who could use any bpi board (not only bpi-r4) for their business projects instead of the home ones.
I am NOT manufacturer with his own business yet, but I am working on it and building my own product right now. Therefore I have important question for @sinovoip : Does the manufacturer wants selling BPI products to home individuals or also for manufacturers for their business? What is the primary target for bpi products?