Feature Request: custom SSH commands + Reboot buttons

[MoneroCrusher]

Hi

It would be great if you could put up a empty box at the top like so:

"Enter your SSH command:"

empty box to enter command

"Dropdown Selection:" Pre-determined SSH: command that the users can create themselves in a database IP: Option "All" or manually select single or pre-defined group of IPs (that the user can define in a database)

RUN# Button

Other small request: Turn on/off machine with a little button in each worker window (like now, but just software/ssh). So there would be 3 buttons: Power off/on button (electricity, like now) additional: Reboot button (normal linux reboot command) Kill-reboot button (in case OS is frozen)

echo 1 > /proc/sys/kernel/sysrq
echo s > /proc/sysrq-trigger
sleep 2
echo b > /proc/sysrq-trigger

Edit: Another feature reqest: In the settings tab, let user choose in "colorization" tab if they want to colorise temperatures and let them define a range and then select the worker they want it to be applied to. Also let them configure a possibility to shutdown/reboot machine if temp of one gpu exceeds pre-defined limit and put it in a logfile (telegram or mail message would be even better, but probably a big effort).

Edit 2: Another request: Add an optional parameter in "colorization" tab, where users can define if they want the worker window to be another color, depending on the pool they input. i.e. if I mine monero on supportxmr.com then I would select "Colorize (color selection)" and input a pool name to search for in the api. So I would select colorize, choose orange and input supportxmr.com This would make the worker window orange. One should be able to define several pools for the same color. if: mining pool = input, then change color according to selection, else: use default Now if I mine Boolberry on another rack, I could select "Colorize", choose blue as color, and now the worker window would be blue. Adding a name would be trivial, too.

In the dashboard total hashrates would be bundled together for different worker colors. in this case there would be a total for orange (xmr) and a total for blue (bbr)

Another general request: Statistics for each worker, where user can define how long to store the data. Interactive graphics would be very nice. But this is porbably a lot of effort.

Another small request: Being able to input the hashrate the rig "should" have, and the one it effectively has. Change worker window color to orange (or a color they want to choose) in case the delta gets too big, which the user can define too. On top of worker window the %of actual hashrate in comparison to "should" hashrate is always displayed. i.e. 96.5%. user can define from which % and down the color should change from default (or colorized, if chosen) to new color of low hash.

[easycryptomining]

Hi,

Your first request is not the aim of ECMGUI : quickly monitor workers not admin them (except for quick reboot from everywhere). It would be a real pain to handle and output correctly the output of the command for each workers (because yes you have to handle that in case there's error, or interactive questions, etc...).

--> so for this sorry but no. There's already lots of tools that do it great : terminator with group, dsh, pssh, cssh, etc... + custom bash alias command.

The reboot request is something i've already think of, but it imply that your ssh user have admin right which is bad especially with ECMGUI as i've made some "quick and dirty" for storing password in the database (in clear text, not crypt).

--> must store password crypt before doing that --> i must think of all imply in term of security and do it the right way --> the sysrq is not that magic, if your OS is frozen, so is your ssh so you won't be able to do it (main aim of sysrq is with key combination on the keyboard)

The temp color is in my todo list EXCEPT for the auto reboot/shutdown (same reasons as the soft reboot button) :

• for each worker 2 new input : warning temp and critical temp
• if cpu or one gpu >= warning temp then orange text
• if cpu or one gpu >= critical temp then red text

--> this will imply another request from people who don't have all the same GPU in the worker, so not the same critical temp which is not little to do...

The worker colors are here to quickly show the worker state (green = ok, orange = pool last see worker more than 1h, red = down). Personally i don't see why you would mine with the same wallet on different pool. Most people i know use one wallet per pool to be reward faster and be able to use the pool tools (stats, api, etc...). Actually in ECMGUI the only pool supported is nanopool and is only here to show the pending balance and to know if the worker send results in the last hour.

--> the best i can imagine to do here is a configurable color per wallet

For the stats i will try not to reinvent the wheel ;)

--> here again there's already lots of tools that do it great : zabbix, munin, nagios, etc...

For the % of expected hashrate, yes good idea, writing it to my todo list.

[MoneroCrusher]

Hey,

1. I already made custom bash scripts now but it would be great if one could command from the dashboard. If you don't want to make it it's time for me to learn php ;)

2. sysrq will always work if you put panic=1 and oops=panic in GRUB file. But it happens often that when 1 GPU crashes in xmr-stak you can't kill the process and rebooting will take like 1 hour. So sysrq is the best solution in that case.

--> the best i can imagine to do here is a configurable color per wallet that would work too and be great!

It would also be nice if we could group certain rigs together.

For the stats i will try not to reinvent the wheel ;)

--> here again there's already lots of tools that do it great : zabbix, munin, nagios, etc... possible/easy to implement in dashboard?

[easycryptomining]

You'll see php is an easy language ;)

I know sysrq always work, as long as you have ssh or physical access to the machine. Here a quick case where ECMGUI will not be able to reboot soft via sysrq :

• Os on an USB key (say 8G)
• xmr-stak running as root (BAD but some guy can do that)
• xmr-stak log file grow and grow for whatever reason --> disk 100% because root used it = no more ssh

==> here the point is not if sysrq is the right answer (it is of course, it save my life multiple time) but that to use it you must be root or sudo. An apps should never be because "With great power comes great responsibility" :) As a sysadmin i don't see myself told people to store password of privileged user in a database (moreover in clear text for now). ==> your point 1 and 2 are link, if you want to make yourself a fork that can execute arbitrary command/script, you'll be able to launch soft reboot script.

Can you give me a used case of grouping rigs that are already groups by wallet ?

Everything is possible, easy depends on backend solution (zabbix as a great api), but please keep in mind that i'm alone working on this project on my free time. As mention in issue #4 ECMGUI is not a light tool (doing some ping, ssh, api call, etc... is very hungry on server/client resources), adding stats even with zabbix api will probably kill the dashboard or the user browser...