[ 386.7 alpha Build(s) ] Testing available build(s)

[L&LD]

Bricked, no.

See my post 273 above.

 

[DJones]

@RMerlin here the logs of aura/led reenabling itself after reboot/scheduled reboot. Maybe it’s amtm led settings interfering.

 

[DJones]

Yep their was a amtm led script. I removed it I don’t remember it affecting aura’s led’s before only the front panel leds. Was previously set to disable led’s at night but when it turns on the front panel it also turns on aura now it seems.

 

[visortgw]

Tried flashing both alpha versions available for my RT-AC68U v4
(RT-AC68U_386.7_alpha1-g82406ddc5d.trx and RT-AC68U_386.7_alpha1-g8ee6420ea6.trx)
to see if this would fix the issue with my 2.4ghz network becomes inoperable after a period of time. Seems others have reported this fix this issue.

When I try to flash says firmware not for that router.

Im gussing they are on 30+ megs instead of 80+ meg that the newer v4 needs or just not built for the Rt-AC68U yet?

Any eta for an alpha that will work on v4?

Separate RT-AC68U v4 image was added to Merlin 386.7 alpha1 builds yesterday afternoon.

 

[greggy101]

In other news, more GPLs came in to merge in the coming days. Hopefully should have them all before the weekend, although things will be more complicated for one of these.

I hope you are not talking about GT-AXE11000... I didn't see any alpha versions for it...

 

[visortgw]

I hope you are not talking about GT-AXE11000... I didn't see any alpha versions for it...

@RMerlin added 45 minutes ago.

 

[Clark Griswald]

FYI signature auto-update working on latest alpha.

 

[RMerlin]

I hope you are not talking about GT-AXE11000... I didn't see any alpha versions for it...

The RT-AC5300 has to be built off a different GPL version due to an issue that it resolved for this model.

 

[L&LD]

Another data point:

I kept the 2x GT-AX6000 routers until the last day I could have possibly returned them. Testing, not just GT-AX6000_386.6_0_nand_squashfs.pkgtb, but also 386.7 Alpha 1 on them for extended periods.

No matter what I did with either firmware, the performance in my environment was markedly worse with this pair of routers in wired (2.5GbE) backhaul mode than the 2x RT-AX86U I have from over 18 months ago today.

Full resets, multiple resets, different types of resets, nothing worked to get the GT-AX6000 to a good/known state.

New SSIDs, no scripts, amtm not enabled, didn't make a difference. Client devices would 'stick' to the main router vs. the wired node. Worse of all (and what finally prompted the return), the wireless throughput was erratic, always changing, and could only occasionally match what the 2x RT-AX86Us could in my environment (no matter what wireless settings were tried over the month).

Initially, kind of surprised at these results, particularly when reading about the many happy users with this model on these forums.

Looking at the differences between my network and others here, the performance of a single GT-AX6000 is superior to the RT-AX86U.

However, when a second (wired, 2.5GbE backhaul) GT-AX6000 router is added as an AiMesh node, the overall wireless network performance tanks by 40% or more.

In comparison, a single RT-AX86U has great performance as I've posted before, When another RT-AX86U is added as a wired node, that performance extends throughout the entire home/property (as expected).

I could have kept the $1K+ routers and hoped that a final 'release' firmware solved the issues. But it was hard to come home to a network that was inferior to the networks I was creating for my customers, and with far cheaper equipment.

 

[capncybo]

Another data point:

I kept the 2x GT-AX6000 routers until the last day I could have possibly returned them. Testing, not just GT-AX6000_386.6_0_nand_squashfs.pkgtb, but also 386.7 Alpha 1 on them for extended periods.

No matter what I did with either firmware, the performance in my environment was markedly worse with this pair of routers in wired (2.5GbE) backhaul mode than the 2x RT-AX86U I have from over 18 months ago today.

Full resets, multiple resets, different types of resets, nothing worked to get the GT-AX6000 to a good/known state.

New SSIDs, no scripts, amtm not enabled, didn't make a difference. Client devices would 'stick' to the main router vs. the wired node. Worse of all (and what finally prompted the return), the wireless throughput was erratic, always changing, and could only occasionally match what the 2x RT-AX86Us could in my environment (no matter what wireless settings were tried over the month).

Initially, kind of surprised at these results, particularly when reading about the many happy users with this model on these forums.

Looking at the differences between my network and others here, the performance of a single GT-AX6000 is superior to the RT-AX86U.

However, when a second (wired, 2.5GbE backhaul) GT-AX6000 router is added as an AiMesh node, the overall wireless network performance tanks by 40% or more.

In comparison, a single RT-AX86U has great performance as I've posted before, When another RT-AX86U is added as a wired node, that performance extends throughout the entire home/property (as expected).

I could have kept the $1K+ routers and hoped that a final 'release' firmware solved the issues. But it was hard to come home to a network that was inferior to the networks I was creating for my customers, and with far cheaper equipment.

OMG, thanks for sharing your real-world testing... I never would have expected such results.
EDIT: rethinking about those numbers it sounds to me like... despite your efforts to reconfigure the 2 GT-AX6000 routers... The Wireless Backhaul was NEVER actually disabled

 

[capncybo]

+ I never meant to insinuate you did anything wrong... I only say this because about 40% hit is what I see when using AiMesh on my home's network.
However in my case... I have a RT-AX86U router & 2 RT-AC-68U nodes.
Okay actually I disliked the AiMesh enough that only One is a Node while the other is an AP... LOL

 

[Nagy Imre]

The RT-AC5300 has to be built off a different GPL version due to an issue that it resolved for this model.

AC88u alpha image does not exist yet. About when will it be available? THX

 

[alienstar]

Performed dirty upgrade on my RT-AC5300 from 386.5_2 to 386.7_alpha1-g07be500548. Everything seems fine after running for 7 and a half hours.

 

[L&LD]

@capncybo, no, trust me. I went far above anything I have done for any client this past month (re: fully resetting and configuring a new network each time).

Even though the 2.5GbE LAN cable (connected to the AiMesh node's 2.5GbE WAN port) was connected, I had to explicitly enable the 2.5GbE port as the preferred backhaul before the node performed as well as the main router.

For a quick recap of what I did (about a dozen times since the first week of May on my network: Before the WAN port on the main router was connected to the ISP (ONT).

• Flashed the RMerlin 386.6_0 (until the 386.7 Alpha 1 was available). Fully reset the router via the reset button, the WPS button method, and via the GUI. This was done on each router, each time, at least once (I did do at least a couple of sets of 10 resets in a row).
• When the RMerlin firmware seemed uncertain (around the third reset, and one set of 10 resets in a row), I flashed the stock Asus firmware and had similar, if not identical results (i.e. using only one main router, the network was fast and stable, if below the total performance of my 2x RT-AX86U setup. Adding the other GT-AX6000 made the network immediately less performant.
• In the next few days/weeks, I did try every combination of my RT-AX86U with the GT-AX6000 in both main router and AiMesh node modes (each). Same results. I had another 2x RT-AX68U that I set up for a customer (initially, at my home), and they too could not keep the performance equal to a single GT-AX6000 (which again, was better than a single RT-AX86U, but not 2x RT-AX86Us in wired (2.5GbE) backhaul mode).
• Removed the 2x QNAP QSW-1105-5T 5-Port Unmanaged 2.5GbE Switches from between the main router location and the AiMesh node location. Same result.
• Tried 2 different (new) cables between the routers (without the 2.5GbE switches in the network), same result.
• I rebooted the node and the main router together (via the AiMesh GUI page) and individually, in all combinations of order, with no improvement.
• I used the RT-AX86U's power adaptors on the GT-AX6000s (and vice-versa), no improvement.
• The 'no improvement' statements above are specifically when a second router/node was added (either another RT-AX86U, GT-AX6000, RT-AX68U, or an older (customer's) RT-AC86U (that last one, I expected to bring the performance down, at least).

During all that time, I left the test GT-AX6000 routers on (and idle, when I wasn't home and not using OpenVPN back to my network). Didn't make a difference if I rebooted them every day, or let them sit for almost 2 weeks straight, the performance never got better with a GT-AX6000 and any other router in the network (as an AiMesh node).

I did more than the above, of course, but that is a good overview.

I was more persistent than I normally would have been (because of the positive results I read about here).

Today, I don't think I can personally recommend the GT-AX6000 if there is even a remote/future possibility of needing an extra node.

I really wanted to make this work (I would have been able to try >1Gbps ISP speeds to see what that is all about (and without needing a specific USB Ethernet 2.5GbE dongle as I would need with my RT-AX86Us.

Hope this is just firmware issues on Asus' side that will be fixed soon.

But, I really don't miss the gaudy styling of the GT-AX6000 either though.

 

[iceorbital]

dirty upgrade on gt-axe11000. So far it hasn't blown up

 

[Morris]

Because it tells you in the first line the name of the process that crashed: "Comm: dcd".

"Broadcom-v8A (DT)" is just the hardware platform of the router and would therefore be the same for every crash on the router regardless of the process that caused it.

got it. Thank you Colin

 

[kernol]

The RT-AC5300 has to be built off a different GPL version due to an issue that it resolved for this model.

Does the same "different GPL version" await the RT-AX86U ?
Is Asus still busy with a permanent patch to the bricking issue - rather than a "work-around" ?

I'm not a coder - but could this bricking issue with the RT-AC86U identified 4+ years ago be similar ?

 

[RMerlin]

Does the same "different GPL version" await the RT-AX86U ?

No. RT-AX86U can still be safely built and flashed provided you use the non-cferom image file.

Is Asus still busy with a permanent patch to the bricking issue - rather than a "work-around" ?

They're testing a fix.

 

[RMerlin]

I'm not a coder - but could this bricking issue with the RT-AC86U identified 4+ years ago be similar ?

These are not discussing update-bricked routers. First one actually had serial output (so it was not bricked), others were having failed wifi radios or bad nand.

 

[octopus]

Beta release is out. we can close this thread.

https://www.snbforums.com/threads/asuswrt-merlin-386-7-beta-is-now-available.79282/

@RMerlin