Hello everyone!
I've created a service for remote rig monitoring. First of all I made it for my own rigs but after time decide bring service to public because it's really useful
It called
RIG cards and available here:
https://rig.cardsBasic functions:- Status of rigs with info for GPUs (temperatures, fan speeds, frequencies etc.) Data for each rig present on its own "card" (as I called it) where you can monitor params just looking to the website
- Notifications to e-mail or Telegram messenger about different events like:
- Rig don't respond / Rig online again
- Temperatures of GPU out of range or back to range (limits can be set for each rig individually)
- Lost GPU on rig
- Rig was rebboted
- Mobile-friendly version of the website, so you can easily check your rigs from smartphone
- Rig rebooting on events like: overheating GPUs, low GPUs temperature, losing GPU on the rig
- Remore rig reboot
- Able to set personal temperature color ranges for rig
Soon:- Hashrate monitoring (I'll plan to add Claymore, EWBF miners first) and notifications which corresponds to it
- Summary card with short status for all rigs, total hashrate, calculated profit, possible troubles
Actions like reboot on different events (for example, reboot when lost GPU) Done!- Other impovements (you can see more detailed tasklist below)
Here is some
screenshots of service main page:
https://i.imgur.com/AIATFbk.pnghttps://i.imgur.com/hCybarz.png
Unfortunately, my status is «newbie» on this forum so my images don't shows yet. Maybe anyone will quote it? thxMy main goal was to make easy in use service which main function is rigs monitoring and not so monstrous as some others.
I'll be happy to get any feedback, suggestions etc.
So, how it works?First you need to register on the
website and download client application that you then will launch on the rig system.
Client app is just Windows Powershell script (like advanced .bat-file), so it's opensource, you can check contents and be sure so there no any viruses or other bad things. It's a text file. Also here is
Virustotal report. Besides it uses library from well-known Open Hardware Monitor to get info from GPU sensors.
View of client window:
https://i.imgur.com/kcyL08h.pngNext you can add new rig from web-interface or during first launch of client app. Choose rig name, set number of GPUs, and temp. limits if wish.
BTW if you already added rig you can download preconfigured client for needed rig and just launch on system.
For launch the client app start the
start_client.bat file (shortcut for this file you can then add to start on windows boot)
https://i.imgur.com/6FkkeVi.pngadding rig from webhttps://i.imgur.com/43tlIko.pngadding rig from client appEvery two minutes client app sends rig data to server. If in 2 min period not received any data from rig, it will marked with "yellow" status (like warning). After next 2-3 minutes if still no rig activity - rig will marked as Offline "red" status and you'll get notification about it through choosen channels in settings.
Also each time when server got data from rig we check given parameters and if something wrong (for example lost GPU or overheating) we also send notification to you (of course if you enabled it).
Temperature limits you can set globally for all rigs (Account - Your Profile - Global settings) or for each rig individually by clicking coghwheel icon for rig you need.
Moreover depends on temp. levels values will be colored, so just taking a look you can check is there everything good with your rigs. You can set you own temperature ranges for each color.
Here is also nice trick with low t° limit: in most cases if something happens with miner on rig (unexpected close, can't connect to pool) GPUs stop working and their temperatures going down. So if find the appropriate value you'll receive notification of low temp and it will mean that GPU don't work normal!
How different statuses looks like:
https://i.imgur.com/qMq2YpQ.pngHere is notifications examples (e-mail and Telegram):
https://i.imgur.com/yJ89HYl.pnghttps://i.imgur.com/wPg7vi4.jpgFor using client app you need to have
.NET Framework 4.5 and
Powershell 5 on your rig systems (Win 10 have them by default, on Win7 usually need to install).
Client has auto update function, but in next releases I add setting to disable auto update if you wish.
How much?It's totally
free for now! No any limitations.
No any fees! And of course it is don't touch your miners.
It just collect info from hardware sensors to show you on web and send nortifications if any.
Known issues and bugs:- It's beta now so I don't except appearing bugs, I'll try to fix asap.
Notification about unexpected rig rebooting doesn't work (not implemented yet, thinking about better algorithm for it) done- Not working with P102-100 or P106-100 — it depends on mode of NVIDIA drivers, solution you can find below in FAQ
- As I see, there no longer any issues here
On first launch of start_client.bat on Windows 10 if you unpacked client.zip by Windows Explorer you'll get Windows SmartScreen warning about unknown app - it's normal, be sure you can launch it without fear (and again link for Virustotal report). There are no warnings if you will unpack zip with Total Commander for example or any other filemanager. As I understood, windows generate such warning every time if you first time start newest .bat which was downloaded from somewhere. And after some time if there no problems, app will be added to their database. You'll get SmartScreen warning only once, on the next launch of app you don't get warning.
To Do list:- Hashrate monitoring and notifications (Claymore and EWBF miners)
Notification for unexpected reboot event doneRig uptime and date/time of launch (for each rig on its card) doneSort & order of rigs' cards done- Sort & order for GPUs on rig
- Summary card with short basic information about all rigs, total hashrate, existing problems and other
- Adding features to Telegram bot (like getting info for rig at request)
- Addition Languages (russian and others)
Actions (like reboot) on different events, possibility to launch bat-file on event- Tasks from server to rig (reboot, shutdown)
- Logo and favicon for website
- Dark/night theme (maybe even posibility to set your own background image)
- Integration with sonoff-tasmota devices through Mosquito servers (MQTT)
FAQHow I can use Telegram bot?Telegram bot is here:
https://t.me/rigcardsbotTo start using you just need to send your secret key to him. Key can be found at Account - Your profile - Profile - Secret key
At this time bot can only send notifications to you, but in future I wish to add some functions
Is it possible to send notification to few Telegram accounts at the same time?First make group in Telegram with few people. Then if you open screen with bot info, at the menu you can choose Add to group. After adding to group send next service message to bot:
/secret your_secretkey_here
Now bot will send notifications to group!
How much I need to pay for using service?It's totally free. But if service come popular I will need to charge for it (to pay for servers and other stuff). But even in this case it will very cheap, 1 USD per month per rig for example (I don't thought seriously about it yet)
GPU NVIDIA P102-100 or P106-100 not showingIt's depends on mode of NVIDIA drivers works. By default it work in
TCC mode and because of it Open Hardware Monitor library that I use to get info from sensors can't get these parameters. Solution is simple - switch drivers to WDDM mode.
According to the official NVIDIA website:
To change the TCC mode, use the NVIDIA SMI utility. This is located by default at C:\Program Files\NVIDIA Corporation\NVSMI. Use the following syntax to change the TCC mode:
nvidia-smi -g {GPU_ID} -dm {0|1}
So you need go to the
C:\Program Files\NVIDIA Corporation\NVSMI directory and from command line run such command for every card pointed it "number" (GPU_ID)
Or (but I'm not sure 'cause I haven't such GPUs) to switch all GPUs on the rig to WDDM mode you can run
Time on rig cards is wrongBy default time on new accounts in UTC timezone. You can set your timezone at Account - Your profile - Timezone
How often data refresh on the page with rig cards?For now data refresh every 30 seconds (but remind, rigs send data every two minutes). If needed you can refresh it immediately with refresh icon on the navbar. After 20 minutes of idle you session will stopped.
How is coloring the GPU temperature values? Is it possible to change it?This coloring is not relating with temperature limits that you set for rigs to get notifications. It just for visual usability. It is possible to set your own temperature ranges for each color.
Default temperature ranges is:
1-35 °C - blue
36-60 - green
61-70 - yellow
71-80 - orange
81 и вышe - red
https://i.imgur.com/KwVhETj.png