v2.8.6 — The Connections page fills in again, and Let's Encrypt stops crying wolf
Five fixes and one diagnostic improvement. Three of the fixes came from the field; two were found
by new checks that now run before a build starts. One of those two had been quietly wrong since
long before this release, on any home network that is not a standard size.
-
Connections — Flow Explorer shows your traffic again. Since v2.8.3 the page had reported
"No active flows in the accelerator cache" however busy the router was. A security check
added to that page's data feed in v2.8.3 was right to be there, but the page itself was never
taught to present the credential the feed had begun asking for — so every request was politely
refused, once a second, and the page drew an empty list. It now sends it. Separately, a refusal
no longer looks like an idle router: if the page cannot read the accelerator for any reason
it now says so, rather than showing an empty table. That ambiguity is why this took a field
report instead of a glance. -
Let's Encrypt stops announcing renewals that did not happen. On the first login after a
restart, the router could announce that your certificate had been renewed and that you were
about to be signed out — with a certificate that had renewed days earlier and was still
perfectly valid. Then it did not sign you out. Two separate faults: the flag the message reads
is set every time the router starts, not only when a certificate is actually renewed; and
nothing in the firmware ever performed the sign-out the message promised. That sign-out was not
cosmetic — the web server only swaps in a new certificate once no one is signed in, so the
missing step was the one that finishes the job. The message now appears only when a certificate
swap is genuinely pending, and it now really does sign you out, which completes the swap. A
renewal that arrives on its own in the background also now arms that swap; previously such a
renewal left the new certificate sitting on disk while the old one stayed in use. -
The lockout protection now covers your whole network. The router keeps a standing rule that
stops its own protection features from cutting you off from it. On the usual home network that
rule was correct, but on a larger one — a /22 or /23, where the network spans more addresses
than the common size — it only covered a quarter or a half of it, and devices outside that
portion had no such protection at all. The rule is now calculated from your actual network and
mask rather than matched against a short list of common shapes. The same fault was found and
fixed in the threat-blocking feature in v2.8.3; this is the same repair applied to the rules
engine, which had been missed. -
Gatekeeper stops reporting a delay that did not happen. On startup Gatekeeper could log that
device controls had been deferred because the network was not ready, while they were in fact
fully applied. Harmless in itself, but it made the message meaningless — a genuine deferral
looked identical to the false one. Fixed. -
Devices left on the removed AdGuard DNS filter are now told. AdGuard was removed as a DNS
filtering choice a few releases ago. Nothing broke — those devices simply resolved names
normally — but any device you had set to it had quietly stopped being filtered while still
appearing to be configured for it, and the setting showed as blank. Such devices are now moved
to Unfiltered, which is what they had actually been all along, and each one is named in the
system log so you can choose a replacement. No device's DNS is silently redirected to a
different provider. -
The diagnostic report answers two more questions on its own. Prompted by a throughput report
that could not be investigated without shell access to someone's router, the report now states
whether the hardware traffic accelerator is actually switched on — previously it described the
accelerator's settings but never said whether it was running, which is the first thing anyone
needs to know when a fast connection is running slowly. It also now reports the automatic PPPoE
redial setting and whether that redial has already happened since the last restart. That setting
is on unless you turn it off, and a blank field previously made it look switched off.
Images & checksums (RT-BE88U)
Two flashable images: + AI Advisor (default) and Standard (noMCP, all AI components compiled out entirely). Flash the *_nand_squashfs.pkgtb via Administration > Firmware Upgrade.
| Variant | File | SHA-256 |
|---|---|---|
| + AI Advisor | RT-BE88U_3006_102.8_Reaper_v2.8.6_nand_squashfs.pkgtb
| 23da74b6d08c8d3cd915f96d44f564402c92eada40c08b57b13b4042d4ee349d
|
| Standard | RT-BE88U_3006_102.8_Reaper_v2.8.6_noMCP_nand_squashfs.pkgtb
| 66e1b93a73f2ace0e941952744c3e2659c03c9f4573c2deee121bd3148a759a3
|
Verify a download against the attached SHA256SUMS-RT-BE88U-Reaper_v2.8.6.txt.
Corresponding source & reproducibility
The RT-BE88U image for v2.8.6 is built from this repository at tag v2.8.6-RT-BE88U: the pinned Asuswrt-Merlin base (3006.102.8-beta2, a7ebfa133a) plus the complete patch series. The tag freezes the exact source that produced it.
- Patches:
patches/(0001-0556) - Provenance record:
provenance/manifest.json - Source tree hash (
release/src/router):6bc355747f14108c6c55b7936426f47a41b391c1-- reproduce bygit am --keep-crof the patches onto the base, thengit rev-parse HEAD:release/src/router. - How to verify:
docs/REPRODUCIBILITY.mdanddocs/SOURCE-AVAILABILITY.md
The auto-attached Source code (zip/tar.gz) asset below is this repository at tag v2.8.6-RT-BE88U (patches + docs).