Antminer R4: full working, but no web-access: Error connection refused!

Helfred

newbie

Activity: 54

Merit: 0

Yes, that can be true: probably it was really 625.

I had several other problems with that model, such as sudden reboots (not temperature related) and sudden stop of hashing.

This was then probably related with the frequency of 625; i do not understand why this is then the frequency factory set from Antminer.

In the end: miner is stable now and on low temps, even with slighty lower hashrate.

What would one want more ?

NotFuzzyWarm

legendary

Activity: 3822

Merit: 2703

Evil beware: We have waffles!

Pretty sure the original freq was ballpark of 625MHz (not 725) with 1 board no doubt being lower than the other. As for speed, I just kept setting the freq higher until the HW error rate started being over 0.002%, for 1 one of my R4's that is 606MHz and another R4 is set to 612.5MHz.

Helfred

newbie

Activity: 54

Merit: 0

As far as i can see now, this embedded image is really stable.

Works slightly under the hashrate of 8.7 promised. Average is 8.5-8.6 . Frequency is 600Mhz.

I think, the image which I bought the R4 with was clocked at 725, but not sure.

What I can say is, that quantity of Hw errors is down by far! was x thousand, now just a few.

What images are others using at which frequency ?

NotFuzzyWarm

legendary

Activity: 3822

Merit: 2703

Evil beware: We have waffles!

No bugs, that version is fine and in 2 respects, better - you should now have fan speed and frequency control back

The only thing you might run into is that now both boards run at the same freq. That means that the maximum speed will be limited to what the weakest board will run at vs autotune usually putting them at different speeds so each boards runs in its sweet spot.

Helfred

newbie

Activity: 54

Merit: 0

@neonox
@NotFuzzyWarm

Gentlemen, you saved me and my miner!

Natasha is now accessible again)))) Thanks to all.

Hardware Version   4.8.1.0
Kernel Version   Linux 3.14.0-xilinx-gf387dab-dirty #38 SMP PREEMPT Fri Jun 17 20:02:51 CST 2016
File System Version   Tue Aug 9 21:00:51 CST 2016
BMminer Version   1.0.0

This is the image now active.

Is it buggy ? Or should I leave it as is ? I have also a backup from the image it was delivered with.

What image is recommended by the users ?
I have
Antminer-R4-all-201612020035-autofreq-user.tar.gz

and the all new
Antminer-R4-all-201704280718-autofreq-user-Update2UBI-NF.tar.gz

Which is the better choice ?

neonox

newbie

Activity: 11

Merit: 0

Quote from: Helfred on May 24, 2017, 12:29:29 AM

Correct ?

Yes, BUT(!) keep in mind:
1) procedure can be different from one version to another
2) you can get a brick until you physically flush box
3) this is just a trick and no guarantee it'll repair your box

Okay, if you read this and understand all risks:
I have no R4 around, so how it work in S5

1) look at the shell script /www/pages/cgi-bin/upgrade_clear.cgi
2) scp downloaded_image.tar.gz to R4
3) tar zxf downloaded_image.tar.gz
if you can see runme.sh - you are lucky
4) sh runme.sh
in other case you should do many wood magic described in upgrade_clear.cgi
after
else
if [ -e /dev/mmcblk0p3 ]; then
...
fi

I uploaded lighthttpd that built for your arch
https://www.dropbox.com/s/p0pypos1y8bqfsq/lighttpd.gz?dl=0

Download it separately on r4 and try to run
/your/folder/lighttpd -v

Helfred

newbie

Activity: 54

Merit: 0

@neonox
@NotFuzzyWarm

Thanks guys for all the help provided.

I do not see the problem solved, but I definitly try the operation recommended by NotFuzzyWarm.
Will be at the miner#s place only by tuesday next week and will keep you informed.

But for general knowledge only:
How would one flash a new image with a ssh session ?

There should be a possibility, since the webinterface probably will only start the command line programs.
Correct ?

Did one try this ?

neonox

newbie

Activity: 11

Merit: 0

Quote from: Helfred on May 22, 2017, 11:09:37 AM

Problem?

Look at the pool stats, I can't see the problem from my end.

Helfred

newbie

Activity: 54

Merit: 0

D****d, i am already out of office!

Hope its not a problem if this process runs the next two days?
Its just killing a lot of pids and restarting them.

i am back in office by wednesday only.

Problem?

neonox

newbie

Activity: 11

Merit: 0

Quote from: Helfred on May 22, 2017, 10:35:29 AM

@neonox
Help! How can this be stopped ?

Use the ps command to get the process id (PID) of this process.
kill -9 (PID)
But, I'm surprised that monitors/supervises lighttpd daemon can stop other process.
https://www.mankier.com/8/lighttpd-angel

Try to reboot, it'll restart all services
reboot

Closing ssh session will not kill daemon:)

Next time please fillow recommendation by @Not Fuzzy warm, it have more sence and less risky than patch application.

Helfred

newbie

Activity: 54

Merit: 0

@neonox

I closed the ssh session. Hope this solved the problem)

I have to leave office now.

Would be great if you support me a little by wednesday ?

Hope on your help.

Helfred

newbie

Activity: 54

Merit: 0

@Not Fuzzy warm

Its worth a try! I am at the moment not at the place as the miner, but this will try definitly!

Helfred

newbie

Activity: 54

Merit: 0

@neonox
Help! How can this be stopped ?

Is there a way to break a command like the Ctrl-C thing ?

NotFuzzyWarm

legendary

Activity: 3822

Merit: 2703

Evil beware: We have waffles!

Have you done the IP Reporter button trick? That will reflash to a backup older version of firmware that is embedded in the R4's.
BE AWARE THAT THIS RESETS ALL SETTINGS TO A FACTORY MODE! YOU WILL NEED TO RENTER ALL THE POOL INFORMATION!

Turn off PSU.
Press and hold in the red IP Reporter button.
While holding the button in, turn on PSU.
Keep pressing the IP Reporter button for 5-7 sec after PSU is turned on.
Release the button. The miner will now flash to backup firmware.
Wait a couple minutes, find the miner on your router (will probably be assigned a new address) and re-enter your pool info.

If all works then get the latest firmware and apply it through the GUI.

Helfred

newbie

Activity: 54

Merit: 0

This command seems to kill a lot of pids and restarting them.

These are a few thousands...?

Command is still running

How can this be stopped ?

neonox

newbie

Activity: 11

Merit: 0

Quote from: NotFuzzyWarm on May 22, 2017, 10:09:31 AM

Quote

And it (flashing to latest firmware) is recommended for all Ant devices as far as I know.

Perhaps recommended by Bitmain, but NOT recommended by us users!
Flashing to latest firmware is always a last-resort move.
As the saying goes, " If it ain't broke... " but in this case, yes, flashing to latest is the next step.

opkg is not Bitmain

https://wiki.openwrt.org/doc/techref/opkg
Upgrading packages is generally NOT recommended for most users, because a typical OpenWrt system stores the base system in a read-only SquashFS partition.

We can try at least there are no other way to solve it.

Quote from: Helfred on May 22, 2017, 10:00:29 AM

How can I flash a firmware through SSH ?

Nice question ) Possible but quite complex.

Can you run
/usr/sbin/lighttpd-angel -f /etc/lighttpd.conf

NotFuzzyWarm

legendary

Activity: 3822

Merit: 2703

Evil beware: We have waffles!

Quote

And it (flashing to latest firmware) is recommended for all Ant devices as far as I know.

Perhaps recommended by Bitmain, but NOT recommended by us users!
Flashing to latest firmware is always a last-resort move.
As the saying goes, " If it ain't broke... " but in this case, yes, flashing to latest is the next step once the miner can be accessed.

Helfred

newbie

Activity: 54

Merit: 0

root@Natasha:~# opkg list-installed | grep [l]ight
lighttpd - 1.4.32-r3.18
lighttpd-module-access - 1.4.32-r3.18
lighttpd-module-accesslog - 1.4.32-r3.18
lighttpd-module-auth - 1.4.32-r3.18
lighttpd-module-cgi - 1.4.32-r3.18
lighttpd-module-dirlisting - 1.4.32-r3.18
lighttpd-module-expire - 1.4.32-r3.18
lighttpd-module-indexfile - 1.4.32-r3.18
lighttpd-module-staticfile - 1.4.32-r3.18

and

root@Natasha:~# opkg status lighttpd
Package: lighttpd
Version: 1.4.32-r3.18
Depends: lighttpd-module-dirlisting, libpcre1 (>= 8.33), lighttpd-module-auth, lighttpd-module-cgi, lighttpd-module-expire, lighttpd-module-access, libc6 (>= 2.18), lighttpd-module-staticfile, lighttpd-module-indexfile, lighttpd-module-accesslog
Recommends: update-rc.d
Provides: lighttpd-systemd
Replaces: lighttpd-systemd
Conflicts: lighttpd-systemd
Status: install user installed
Architecture: armv7ahf-vfp-neon
Conffiles:
/etc/lighttpd.conf a7a5b79321fbb063fe2aad1fa350ffe1
Installed-Time: 1468402192

How can I flash a firmware through SSH ?

neonox

newbie

Activity: 11

Merit: 0

Quote from: Helfred on May 22, 2017, 09:41:11 AM

What are these magic words mean ?)

It provides some basic information about this binary.
Show me your:
opkg list-installed | grep [l]ight
and
opkg status lighttpd

Ok, seems the better way for you is just reflash Natasha with a newer firmware image. And it is recommended for all Ant devices as far as I know.

Helfred

newbie

Activity: 54

Merit: 0

root@Natasha:~# stat /usr/sbin/lighttpd
File: /usr/sbin/lighttpd
Size: 117720 Blocks: 230 IO Block: 4096 regular file
Device: 1f01h/7937d Inode: 1179 Links: 1
Access: (0755/-rwxr-xr-x) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2016-07-13 09:28:56.000000000
Modify: 2016-07-13 09:28:56.000000000
Change: 2016-07-13 09:28:56.000000000

What are these magic words mean ?)

root@Natasha:~# which lighttpd
/usr/sbin/lighttpd

Topic: Antminer R4: full working, but no web-access: Error connection refused! (Read 1656 times)