Connection Type: DHCP
Ignore WAN DNS: IPv4 DNS 0/IPv4 DNS 1
Now able to set just two Static DNS. Ignore WAN DNS Works as intended.
Syslog is the same as DNS set. no-resolv/strict-order no longer needed.
Browser: Firefox Version 78.0.2 (64-bit)
Router Model: Linksys WRT3200ACM
Previous Version: r41379 (aka r41380) Default Build
Firmware Version: DD-WRT v3.0-r44048 std (08/02/20) Webflashed from build r41379
Kernel Version: Linux 4.9.231 #1979 SMP Sun Aug 2 03:35:09 +03 2020 armv7l
OpenVPN Client: (NordVPN Server) OK
5GHz 40MHz Mixed 36 + 40 5180MHz WPA2 CCMP-128 (AES) OK
(Received RX OK, no error) (Transmitted TX OK, no error)
2.4GHz 20MHz Mixed 11-2462MHz WPA2 CCMP-128 (AES) OK
(Received RX OK, no error) (Transmitted TX OK, no error)
Wireless Physical Interface ath2 [2.4GHz/5 GHz/802.11ac] - SD8887 802.11ac
Disabled. Shows on Interface OK
Load Average: 0.00, 0.00, 0.00 0% OK
Temperatures: CPU 56.0 °C / WL0 32.8 °C / WL1 33.9 °C OK (USB cooling fan)
Clients: 2 PS3s, 2 PS4s, 2 Firesticks, 1 Firebox, 3 Android phones, 2 Laptops, 1 Desktop
Status: Webflash no reset, Uptime 1 day, 14:34, QoS FQ_CODEL MAC Priority, no keep alive, or reboot needed OK.
*Using (Speedtest) Android App will cause WiFi error's.
Last edited by Justanotherbrokenrouter on Wed Aug 05, 2020 17:59; edited 1 time in total
Router/Version: WRT32x
File/Kernel: Linux 4.9.231 #1979 SMP Sun Aug 2 03:35:09 +03 2020 armv7l
Previous/Reset: 07-21-2020-r43886
Mode/Status: AP mode only, no DHCP (DHCP forwarder enable), everything works fine, uptime 2 days, 20:41
Issues/Errors: Load average still 50% load when idle, ESP8266 needs disable WMM support then it is able to connect WRT32x.
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Thu Aug 06, 2020 14:24 Post subject:
here from my wrt3200acm
no load issues
straight wifi client operation
root@PS4:~# uptime
16:23:11 up 1 day, 4:00, load average: 0.00, 0.00, 0.00
root@PS4:~# nvram get DD_BOARD
Linksys WRT3200ACM _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Thu Aug 06, 2020 14:30 Post subject:
PavelVD wrote:
Your comments are fair, I'm trying to get better.
Since SAMBA has been updated, I've really had problems. Every time I experience a new build, but I return to r42054. I carefully follow the messages on the forum and read the tips, in particular tried this:
https://forum.dd-wrt.com/phpBB2/viewtopic.php?p=1212244#1212244 (Tried other protocols, the result does not change.)
There's a phrase:
SurprisedItWorks wrote:
I gather that Samba issues in recent builds is due to a move a small number of months ago to Samba code embedded in the linux kernel. It's taken a while to get the kinks out. Maybe some are still in there.
That's why I report problems in the thread so that developers can see them, and wait patiently.
Each build contains a message:
kernel-panic69 wrote:
Issues:
• There may be remaining issues for Samba (for example NTFS), with frequent updates.
I tried to change the markings of the discs, but it didn't help me either.
I'm not doing anything out of the ordinary. I work on a laptop connected by LAN. If necessary, I turn off the wire and switch to Wi-Fi, that's where it all starts.
I'm trying to do it specifically now for the test. Preliminarily reset the reset button and reintroduced all settings.
Router restarted and wrote this text; Connection - LAN.
As soon as LAN was disabled, my PC lost its network drive. After a few seconds, the drive became available and I made an attempt to copy the file on the PC. Copying failed, here's what appeared in the log:
Code:
Aug 4 19:37:55 LinkSYS daemon.info dnsmasq-dhcp[2510]: DHCPREQUEST(br0) 192.168.1.121 20:0d:b0:41:c6:a9
Aug 4 19:37:55 LinkSYS daemon.info dnsmasq-dhcp[2510]: DHCPACK(br0) 192.168.1.121 20:0d:b0:41:c6:a9 MyLenovo
Aug 4 16:37:58 LinkSYS kern.debug kernel: [ 4947.364491] ieee80211 phy0: Mac80211 start BA 20:0d:b0:41:c6:a9
Aug 4 16:38:04 LinkSYS authpriv.info dropbear[4305]: Child connection from 192.168.1.121:57758
Aug 4 16:38:05 LinkSYS authpriv.notice dropbear[4305]: Password auth succeeded for 'root' from 192.168.1.121:57758
Aug 4 16:41:01 LinkSYS daemon.err httpd[1560]: httpd : Request Error Code 408: No request appeared within a reasonable time period.
Aug 4 16:41:54 LinkSYS kern.err kernel: [ 5183.281481] ksmbd: ksmbd_conn_write:214: Failed to send message: -11
Aug 4 16:41:59 LinkSYS kern.err kernel: [ 5188.321487] ksmbd: ksmbd_conn_write:214: Failed to send message: -11
Aug 4 16:42:04 LinkSYS kern.err kernel: [ 5193.371468] ksmbd: ksmbd_conn_write:214: Failed to send message: -11
Further work on Wi-Fi (5 GHz) became impossible, I connected the wire.
working daily with samba. consider to set protocol to minimum of smb2 if you want to use smb3. windows still requires this minimum support unlike linux. beside this i havent seen any issues on my dd-wrt based nas so far.
and about your 5 ghz issues. i see no problem description in your comment _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Thu Aug 06, 2020 14:31 Post subject:
Argenis wrote:
Upgraded 3200acm from 23 July build, and 32x from 32 July build.
3200ACM had to re-setup the password for one of the wireless networks, after that all good, no issues.
32x still has the 50% load issue (with close to 0% cpu usage).
the load is the cpu usage. cpu usage cannot be 0 if load is 50%. 2.0 load means both cores are fully used. so 25% of your device is current fully loaded. (vpn, bandwidth, ddos attacks ) _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
the load is the cpu usage. cpu usage cannot be 0 if load is 50%. 2.0 load means both cores are fully used. so 25% of your device is current fully loaded. (vpn, bandwidth, ddos attacks )
Precisely. This is why when we see a never varying/steady 1.0 load in all three positions with router basically idling, (CPU: 0.0% using top command in putty) we know something is wrong. In my case, an upgrade of my 3200acm to any recent build, or resetting any recent build to factory defaults, or nvram erase && reboot will cause this, even without internet connection of any kind.
Restoring a "working" backup, (install a "pre 1.0 load issue" firmware version, set all settings, upgrade to any newer beta WITHOUT resetting, then backup) will allow for proper functioning (load minimums below 1.0 possible) - the ath2 visible/not visible thing may provide a clue, because when ath2 is visible, load display functions properly, when ath2 is "gone," load never displays less than 1.0 (reboot of router "clears" load, starts out below 1.0, but once it rises above 1.0, it never goes lower)
Joined: 30 May 2017 Posts: 582 Location: Rural Manitoba
Posted: Thu Aug 06, 2020 21:45 Post subject:
jd8121 wrote:
BrainSlayer wrote:
the load is the cpu usage. cpu usage cannot be 0 if load is 50%. 2.0 load means both cores are fully used. so 25% of your device is current fully loaded. (vpn, bandwidth, ddos attacks )
Precisely. This is why when we see a never varying/steady 1.0 load in all three positions with router basically idling, (CPU: 0.0% using top command in putty) we know something is wrong. In my case, an upgrade of my 3200acm to any recent build, or resetting any recent build to factory defaults, or nvram erase && reboot will cause this, even without internet connection of any kind.
Restoring a "working" backup, (install a "pre 1.0 load issue" firmware version, set all settings, upgrade to any newer beta WITHOUT resetting, then backup) will allow for proper functioning (load minimums below 1.0 possible) - the ath2 visible/not visible thing may provide a clue, because when ath2 is visible, load display functions properly, when ath2 is "gone," load never displays less than 1.0 (reboot of router "clears" load, starts out below 1.0, but once it rises above 1.0, it never goes lower)
Observation on my part; If Ath2 is not visible on newer builds it is an indication it is active (enabled). I found many months ago that Ath2 must be disabled before any configuration is done with new builds. Failure to do so will default Ath2 enabled and on the first reboot it will disappear from the GUI and Acrylic WiFi confirms that it is active. _________________ Starlink & DSL -> TPLink TL-R470T+
->
WRT3200acm Master WDS 5GHz 80Mhz CH 100 (+6) r55819
Ath1 2.4Ghz Disabled
99 Static Leases
ExpressVPN
WRT3200acm r55819 WDS Station 5Ghz
Ath1 AP N/G Mixed Channel 11 HT40
WRT1900Ac V1 5Ghz r55819 WDS Station
(Defective, no 2.4Ghz but 5Ghz works great)
WRT1900AC V1 5Ghz AC 80Mhz WDS Station r55819
2.4Ghz AP Ch1 HT20 Mixed
WRT1900ACS SPARE r54914
WRT1900AC SPARE r54914
WRT1900AC V1 5Ghz AC 80Mhz WDS-AP r55819
2.4Ghz AP Ch1 HT20 Mixed
WRT54G DD-WRT v3.0-r37305 micro AP CH 6 Mixed - Not in use
working daily with samba. consider to set protocol to minimum of smb2 if you want to use smb3. windows still requires this minimum support unlike linux. beside this i havent seen any issues on my dd-wrt based nas so far.
and about your 5 ghz issues. i see no problem description in your comment
Perhaps I formulated something incorrectly, or this is a problem with online translation, I will try to rephrase:
Wi-Fi (5 GHz) works well without using an SMB disk. In the above piece of the log, you can see that the errors appear after about 1 minute (the time to reconnect the SMB disk when changing LAN to WLAN). Those. while I work on a laptop over the wire, I do not experience any special problems, however, if I need to change the connection,… butt ass. (Sorry for my "French".) And, regardless of the type of connection, deleting a file on a network drive with Windows Explorer (caused by the Win+E combination) received a reboot of the router, after which a wire must be attached to the laptop - otherwise you cannot get access to the network drive ... (To avoid such a situation, I had to learn how to delete files without using Explorer. But not everyone in my family can do this.) After such a restart, there are no entries in the "old" log that could shed light on the problems and present them here - on the forum ... (Each time the router starts, a new log file is created, the old one is saved.)
When deleting a file on a "good" PC, you could see this picture for 5-10 minutes:
A small digression: in the last sentences, verbs in the past tense were used all the time - Why? I have at my disposal three PCs with Windows 10 Home on board, the most buggy one has a 32-bit system. Yesterday I could not stand the torment, reinstalled Windows - installed a 64-bit system - the laptop allowed it. And ... Hallelujah! - The problems are gone! I can't say with certainty that everything is fine in general - too little time has passed for such a statement. At least right now I am writing this text and have no problems.
Maybe other forum readers experiencing problems with SMB will also pay attention to the bitness of their systems? (Maybe the version of the systems matters? - I don't understand how this can affect the router.)
Thank you Mr. SurprisedItWorks, I will certainly apply your advice about SMB 2. _________________ Linksys WRT1900ACSv2
Automatically adjustable temperature, always within the range of 59-68°С.
Last edited by PavelVD on Thu Aug 06, 2020 22:23; edited 1 time in total
here from my wrt3200acm
no load issues
straight wifi client operation
root@PS4:~# uptime
16:23:11 up 1 day, 4:00, load average: 0.00, 0.00, 0.00
root@PS4:~# nvram get DD_BOARD
Linksys WRT3200ACM
You're correct, my 3200ACM has no issues with the build, whatsoever, and I will continue using it and updating it to the latest. I can see why the 32x got discontinued quickly and gets no love from the manufacturer, so I'll just retire it. _________________ Router: Linksys WRT3200ACM WLAN0 and 1 have same SSID
88W8964 802.11ac WLAN0 Mode AP VHT80 80MHz Mixed Mode Channel and Extension Channel Auto Extension LL-6
88W8964 802.11ac WLAN1 Mode AP 20 MHz Mixed Mode Channel Auto
SD8887 802.11ac disabled but visible on GUI and CLI
TX Power 18 dBm
Antenna Gain 0 dBi
U-APSD (Automatic Power Save)Enabled
Protection Mode None
RTS Threshold Disabled
Short Preamble Disabled
Short GI Enabled
Single User Beamforming Enabled
Multi User Beamforming Enabled
AP Isolation Disabled
Beacon Interval 100
DTIM Interval 2
WMM Support Enabled
Radar Detection Disabled
ScanList default
Sensitivity Range (ACK Timing) 500 (Default: 500 meters)
Max Associated Clients 256 (Default: 256 Clients)
Minimum Signal for authenticate -128
Minimum Signal for connection -128
Poll Time for signal lookup 10
Amount of allowed low signals 3
Wireless security is WPA2 Personal CCMP-128 only
QAM256 is on
Observation on my part; If Ath2 is not visible on newer builds it is an indication it is active (enabled). I found many months ago that Ath2 must be disabled before any configuration is done with new builds. Failure to do so will default Ath2 enabled and on the first reboot it will disappear from the GUI and Acrylic WiFi confirms that it is active.
I do remember reading the posts about that when troubleshooting. I've only accomplished ath2 disabled/visible if I upgrade and don't reset. Would be very happy to find it there for me to disable on first "with reset" boot! Will check for ath2 *first* when I upgrade again/next beta (I usually apply other settings first, before wireless, when I upgrade with reset, and even without reboot, ath2 is not there - I presumed that was because even the router's initial post upgrade boot was using default of enabled)
Router/Version: WRT32X
Firmware: DD-WRT v3.0-r44048 std (08/02/20) (previous 07-23-2020-r43904)
Kernel: stock
Mode: Gateway, 2.4G and 5G wifi disabled, QOS CAKE, upnp
Reset: No from last build 07-23-2020-r43904
Did have the CPU usage moving from 40-50% (attached) now stuck on 50% according to the gui. Which doesn't really matter to me anyway. The build seems snappy and hopefully stable, Thanks BS and all involved.
Switched the SMB protocols to 2.02/3.00 log filled with view messages:
Code:
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.854491] ------------[ cut here ]------------
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.859188] WARNING: CPU: 0 PID: 3727 at /home/seg/DEV/mvebu/src/router/smbd-next/smbd/buffer_pool.c:364 ksmbd_free_work_struct+0x40/0xa4 [ksmbd]
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.872350] Modules linked in: ksmbd ext4 jbd2 mbcache crc32_generic ledtrig_usbport pwrseq_simple pwrseq_emmc ahci_mvebu ahci_platform ahci libahci_platform libahci ehci_orion sata_mv usb_storage sr_mod cdrom sd_mod xhci_pla
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.929924] CPU: 0 PID: 3727 Comm: kworker/0:0 Tainted: G W 4.9.231 #1979
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.937873] Hardware name: Marvell Armada 380/385 (Device Tree)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.943828] Workqueue: ksmbd-io handle_ksmbd_work [ksmbd]
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.949269] [<c011356c>] (unwind_backtrace) from [<c010ee54>] (show_stack+0x10/0x14)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.957048] [<c010ee54>] (show_stack) from [<c03145b0>] (dump_stack+0x90/0xa4)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.964304] [<c03145b0>] (dump_stack) from [<c01b63b0>] (__warn+0xe4/0x100)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.971297] [<c01b63b0>] (__warn) from [<c0129cf0>] (warn_slowpath_null+0x20/0x28)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.978907] [<c0129cf0>] (warn_slowpath_null) from [<bf5a6140>] (ksmbd_free_work_struct+0x40/0xa4 [ksmbd])
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.988616] [<bf5a6140>] (ksmbd_free_work_struct [ksmbd]) from [<bf59ca5c>] (__smb2_oplock_break_noti+0x298/0x2c0 [ksmbd])
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7983.999719] [<bf59ca5c>] (__smb2_oplock_break_noti [ksmbd]) from [<bf59d28c>] (__smb2_lease_break_noti+0x620/0x828 [ksmbd])
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.010909] [<bf59d28c>] (__smb2_lease_break_noti [ksmbd]) from [<bf59e9f8>] (smb_break_all_levII_oplock+0x1a8/0x2ac [ksmbd])
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.022274] [<bf59e9f8>] (smb_break_all_levII_oplock [ksmbd]) from [<bf5b2b48>] (smb2_lock+0xde8/0x1110 [ksmbd])
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.032505] [<bf5b2b48>] (smb2_lock [ksmbd]) from [<bf5a4598>] (handle_ksmbd_work+0x2c0/0x474 [ksmbd])
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.041860] [<bf5a4598>] (handle_ksmbd_work [ksmbd]) from [<c01430f4>] (process_one_work+0x284/0x408)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.051119] [<c01430f4>] (process_one_work) from [<c0143630>] (worker_thread+0x3b8/0x5cc)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.059332] [<c0143630>] (worker_thread) from [<c0148550>] (kthread+0x150/0x158)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.066760] [<c0148550>] (kthread) from [<c010ba80>] (ret_from_fork+0x14/0x34)
Aug 7 13:01:03 LinkSYS kern.warn kernel: [ 7984.074038] ---[ end trace 7a08bd5d4b4c118f ]---
Lots of messages.
Customer SMB (only on the net at the time of verification, but there are NFS customers): Windows 10 Home based x64, Version 10.0.19041 Build 19041.
At the same time, he did not do any manipulations with the files, typed the text in the application OneNote Office 2010, the notebook of which lies on a shattered disk.
Switched the SMB protocols to 3.00/3.11 - the log is clean. _________________ Linksys WRT1900ACSv2
Automatically adjustable temperature, always within the range of 59-68°С.
Joined: 04 Aug 2018 Posts: 1447 Location: Appalachian mountains, USA
Posted: Sun Aug 09, 2020 16:49 Post subject:
Router/Version: WRT1900ACSv2
File/Kernel: ddwrt-linksys-wrt1900acsv2-webflash.bin
Previous/Reset: flashed 44048 from 42926 from CLI with no reset
Mode/Status:
Up max 3.5 days so far, as I reboot when changing parameters
Typically 15-20 clients, ISP 200/10 DHCP connect, light load with two TV streams (bypassing VPN) max
Gateway mode,
2.4GHz NG-Mixed HT40 SGI short preamble,
5GHz AC/N-Mixed VHT80 SGI short preamble and single-mode beamforming,
three 2.4GHz VAPs,
LANs split into two VLANs
OpenVPN client/PBR on AirVPN (referral link), 61 Mbps down,
TWO wireguard clients/PBR on different AzireVPN (vanilla link) servers, 181 Mbps / 108 Mbps down from the less/more loaded one,
Edit: 215.5/11.5 Mbps via 5GHz wifi and Azire wireguard (server 1 Kmi or so away) over a 200/10 ISP service:
USB/NAS large ext2 flash drive SMB (min/max 3.0/3.11) mapped in MacOS and Windows-10 and mounted in Fedora (fc29) Linux
DNSCrypt (built-in dnscrypt-proxy) from two providers,
"addn-hosts=..." DNS-sinkhole adblocker (90K FQDNs),
DNS autorouted through OpenVPN client when ping tests permit
Issues/Errors:
The color styles in "Router GUI Style" in GUI>Administration>Management don't work (Fedora fc29 linux Firefox 70), at least not in Preview. This was also the case in 42926 and 43904. They did work at some point within the past three years (also Fedora linux Firefox), but I don't remember what builds.
This build also still has (as did my last builds: 42926, 43904) an OpenVPN firewall error, in that the /tmp/openvpncl/route-up.sh line
iptables -t raw -I PREROUTING ! -i $dev -d $ifconfig_local/$route_netmask_1 -j DROP
(and the corresponding -D route-down.sh line) fails and creates an error in the vpn log because $route_netmask_1 is not defined. The intended firewall rule is not created. Replacing the above (and similarly fixing corresponding -D route-down.sh line) with
iptables -t raw -I PREROUTING ! -i $dev -d $ifconfig_local/$ifconfig_netmask -j DROP
fixes it. It's probably the least-important OpenVPN-client firewall rule, but perfectionist that I am, I use a workaround: create an up.sh that edits the route-*.sh files. I can't create it before openvpn is run, however, so openvpn crashes and must be restarted. Awkward and kind of in the "do not try this at home" category.
Edit: The cause of this problem has been identified, with help from egc, so it'll get fixed eventually.
_________________ 2x Netgear XR500 and 3x Linksys WRT1900ACSv2 on 53544: VLANs, VAPs, NAS, station mode, OpenVPN client (AirVPN), wireguard server (AirVPN port forward) and clients (AzireVPN, AirVPN, private), 3 DNSCrypt providers via VPN.
Joined: 04 Aug 2018 Posts: 1447 Location: Appalachian mountains, USA
Posted: Sun Aug 09, 2020 19:03 Post subject:
egc wrote:
Works for me at least for my routers:
Code:
root@EA6900:~# iptables -vnL -t raw
Chain PREROUTING (policy ACCEPT 3250K packets, 1356M bytes)
pkts bytes target prot opt in out source destination
0 0 DROP 0 -- !tun1 * 0.0.0.0/0 10.116.1.6
Those are environment variables and can only be called from within OpenVPN and will not work outside the OpenVPN environment
When you make your own scripts you have to take that into account
I don't call them from anywhere unusual. I work the fix by having up.sh edit the two route-*.sh files, which are then called from OpenVPN in the usual way at the usual time. All that works fine. But the original problem occurs, of course, before I have touched the setup at all. In that case, I get the errors in the vpn log, and iptables -t raw -vnL PREROUTING shows nothing at all (or no tun1 rule if I have set up wireguard). I certainly can't explain why it happens on all five of my family's routers and not on yours!
To other Marvell people running the OpenVPN client: (when you have the CVE mitigation patch enabled -- thanks @egc -- for the OpenVPN client) does iptables -t raw -vnL PREROUTING show you a rule with !tun1 in it? _________________ 2x Netgear XR500 and 3x Linksys WRT1900ACSv2 on 53544: VLANs, VAPs, NAS, station mode, OpenVPN client (AirVPN), wireguard server (AirVPN port forward) and clients (AzireVPN, AirVPN, private), 3 DNSCrypt providers via VPN.
)