pi node container consensus issue

[alexmihai212]

Troubleshooting Diagnostics Info Container Consensus container enabled: No -HOW to ENABLE? Node switch on: Yes PortsChecker container: running Consensus container: absent **-HOW to ENABLE?

**

Other Pi Node version number: 0.3.8 Operating system: WINDOWS 10.0.19042 (Microsoft Windows 10 Enterprise) Using Docker Toolbox: No Node installation date: 2021-01-14 Availability (up to 90 days): 37.18% from 2021-01-14 to 2021-01-19

Back

[ihatejam]

Keep running the node listener container until your availability stats are higher - this will take a while. Once you approach or exceed 90% availability, you are more likely to be selected to run the consensus container. Once selected, the consensus container will be auto-deployed to your node setup.

[prexone]

I have a similar issue. My consensus container keeps switching from absent to error. 82.97% Availability. Is my solution different?

https://share.getcloudapp.com/o0uEBZe8

[dgagvgeg]

ahh having the same issue, just keep it up and running I guess lol

[ihatejam]

Correct - just leave it running - the bit where the status alternates between absent and error was added with 0.4.x at the end of last week. Once you're selected to be part of the node test, a consensus container will be sent down to you and you'll then see "Consensus container enabled: Yes" and the "Consensus container:" part will temporarily alternate between creating/absent and eventually settle on "Consensus container: running" (assuming of course that ports are correctly forwarded, etc.).

[shafay07]

After the update to version v0.4.2 (proton/-), i had to do the port forwarding again. Everything is ok i keep the node running but my availability time is reducing everyday, i have the same notification for consensus container (switching from error to absent). any fixes?

[theoptips]

To reach 90% + availability, does that mean we need to keep the Pi desktop app running for more than 21.6 hours a day? Thanks. My consensus container is OFF. @ihatejam thanks for the insight. I didn't know that. Their doc also kind of infer that only Super Node runs consensus.

[Hany6656]

لا استطيع ان اجعل الاتصالات الواردة اكثر من صفر . ماذا أفعل ؟

[BlackBeardo]

I am already synced.

[AMZNG1]

I am already synced.

How long it took u to be synced?

[AlfaB16]

Hi, Good morning. I have tried to run the Pi Node on my pc running windows 10 pro (Licensed) but till now I haven't managed to get the 10 required ports (31400-31409) to open although I have isolated my computer from the rest of the devices on my home network by providing my PC with a static IP address and have access to the internet as well. Docker is also running. I am using a Dual band Router ( TP-Link Acher C20 ) If you do have the time please let me know where I have made a mistake.

Thanks and regards.

[willemPK]

Hello everyone,

I've got some problem with my ports 31400-31409. I did everything that was said to be done in "https://www.youtube.com/watch?v=QBDQeNVRNak" But every time I scan, it shows me half of the ten ports open and always different ports.

[Jepeto78]

Hello everyone,

I've got some problem with my ports 31400-31409. I did everything that was said to be done in "https://www.youtube.com/watch?v=QBDQeNVRNak" But every time I scan, it shows me half of the ten ports open and always different ports.

Got exactly the same issue (running WS2019 with linux container enabled, firewall configured, external port scan OK) : single action was to reboot => after all, all ports were green.

Now Consensus Container status blinking between absent & creating (without any logs ...)

[ihatejam]

Hello everyone,

I've got some problem with my ports 31400-31409. I did everything that was said to be done in "https://www.youtube.com/watch?v=QBDQeNVRNak" But every time I scan, it shows me half of the ten ports open and always different ports.

@willemPK - you should be good to go - the built-in portscan check is temperamental. You can sanity-check your ports at https://www.ipfingerprints.com/portscan.php

[ihatejam]

Got exactly the same issue (running WS2019 with linux container enabled, firewall configured, external port scan OK) : single action was to reboot => after all, all ports were green.

Now Consensus Container status blinking between absent & creating (without any logs ...)

@Jepeto78 As far as I'm aware nobody has gotten it to work with windows server - even with linux containers enabled in docker. Best thing to do is spin up an installation of windows 10 home/pro (at least build 2004) with Docker Desktop. Check out the requirements at node.minepi.com

[Jepeto78]

Got exactly the same issue (running WS2019 with linux container enabled, firewall configured, external port scan OK) : single action was to reboot => after all, all ports were green. Now Consensus Container status blinking between absent & creating (without any logs ...)

@Jepeto78 As far as I'm aware nobody has gotten it to work with windows server - even with linux containers enabled in docker. Best thing to do is spin up an installation of windows 10 home/pro (at least build 2004) with Docker Desktop. Check out the requirements at node.minepi.com

Or maybe someone could have that thing fixed ?

Handler for POST /v1.40/containers/create returned error: invalid volume specification: 'C:\Users\Jeep\AppData\Roaming\Pi Network\docker_volumes\stellar:/opt/stellar': invalid mount config for type "bind": bind source path does not exist: c:\users\jeep\appdata\roaming\pi network\docker_volumes\stellar

[Jepeto78]

Got exactly the same issue (running WS2019 with linux container enabled, firewall configured, external port scan OK) : single action was to reboot => after all, all ports were green. Now Consensus Container status blinking between absent & creating (without any logs ...)

@Jepeto78 As far as I'm aware nobody has gotten it to work with windows server - even with linux containers enabled in docker. Best thing to do is spin up an installation of windows 10 home/pro (at least build 2004) with Docker Desktop. Check out the requirements at node.minepi.com

Or maybe someone could have that thing fixed ?

Handler for POST /v1.40/containers/create returned error: invalid volume specification: 'C:\Users\Jeep\AppData\Roaming\Pi Network\docker_volumes\stellar:/opt/stellar': invalid mount config for type "bind": bind source path does not exist: c:\users\jeep\appdata\roaming\pi network\docker_volumes\stellar

Achieved to make it start (at least) on WS2k19 while:

• excluding "appdata\roaming\pi network\docker_volumes" from windows defender policies
• creating manually the 3 folders which were reported as non existing by docker bind docker_volumes\stellar, docker_volumes\supervisor_logs, docker_volumes\history Container has then started, node was synching but you may be right:

Handler for HEAD /v1.40/containers/pi-consensus/archive returned error: filesystem operations against a running Hyper-V container are not supported Handler for PUT /v1.40/containers/pi-consensus/archive returned error: filesystem operations against a running Hyper-V container are not supported

too bad :)

[ihatejam]

i wasn't too clear in my previous message -- it's not a Pi node problem -- afaik -- it's more a case of running a docker linux container "properly" with full fs perms on windows server that has never really worked 100% -- Even if you took the time to switch on the nested virtualization stuff and flipped the zillions of hyper-v switches (many via Powershell). MS and Docker tried to make it work better over the last couple of years - and it has become better, but some bits still suck.

Using a server class OS (presumably running on server-class hardware) is an overkill for an SCP-based node though. A semi decent wintel box running win10 is sufficient.

[dgennetten]

I've been running node successfully on one PC. All subsequent attempts to install and run on other macs or PCs result in the same error: the consensus container doesn't get loaded. Even with all ports open (which wasn't the case with my one successful install). Is this just a matter of waiting?

Fresh install/updates of everything. Maybe it's because I'm allowed only to get one node running?

[kulveersingh]

How much is the reward for running the node?

[knight0zh]

Hi there Installed about two days ago, but It can not show installation day and same "ABSENT <>ERROR" problem I do have ... does anybody know what is this, why is this and does it worth to run this node ( which full with unknown errors and looks like a real scam)?

Got exactly the same issue, Is your node normal now？

[steeveee100]

Can you share the data.. troubleshooting + results from command docker ps -a and wsl -l -v from command prompt?

[steeveee100]

The reason I ask this is to help you, same as I got helped recently. But it depends on your results.

[nathusius]

Hi there Installed about two days ago, but It can not show installation day and same "ABSENT <>ERROR" problem I do have ... does anybody know what is this, why is this and does it worth to run this node ( which full with unknown errors and looks like a real scam)?

Got exactly the same issue, Is your node normal now？

The absent <> error for the consensus container is normal behaviour, until you get picked to run the full node, so just leave it running for the max time possible! If/when you do get picked, the consensus container will be downloaded to you and start running and the ports checker container with show "exited"; at that point it will also show the installtion date for the node. You have to be patient, two days doesn't sound like nearly enough time to get selected, more like 2 weeks or more is needed! And if you think it's a scam then just delete it from your machine!

[ihatejam]

FYI - There's a newer version 0.4.4 available on node.minepi.com - give it a try Previous versions did not give the ability to switch it off the API service -- after installing 0.4.4, the API service will be switched off by default which saves disk space.

[ If you already got the node going on 0.4.3 or earlier, you can go back to reclaim disk space, by doing a "remove all blockchain data" on the troubleshooting page, and then it'll kick off a much leaner existence (give it a few minutes to sync back up the latest block -- for me, it seemed to stick at block 1 for a good few minutes and then once it got going ramped up to the latest block in almost no time). I went from a 40GB docker volume down to 2.9GB ]

[nathusius]

Thanks, I just installed 0.4.4, also found that it sat on block 1 for quite a while (30 mins maybe?) then quickly caught up; I saw that there were some incoming connections, even while it was still on block 1, so that seemed a little bit surprising?! Where can I see the docker disk space used? (just curious really!) Any thoughts on what is a "normal" number of incoming/outgoing connections? I generally sit at 8 outgoing (maximum), but only a few incoming, even though the limit is higher at 64 (1~3 is typical, quite often 0, sometimes a little higher at 4 or 5, but always less than 8); again really just curious to know if everything is OK! Numbers have gone up while I've been typing, now sitting quite steady at 5 or 6 incoming!

[anon-netizen]

[](

Is this running correctly?

[steeveee100]

Looks good! But do you ever get any incoming connections?

[anon-netizen]

Looks good! But do you ever get any incoming connections?

Not so far! That was my concern. Afaik, the ports are all correctly configured.

[steeveee100]

Ok. I'm not an expert, but does State ever say "Synced!"? And Latest Block change to "a few seconds ago"?

What do you get when you hit the back button? What is your report below the Node On button?

Also is your device on 24/7?

[anon-netizen]

Ok. I'm not an expert, but does State ever say "Synced!"? And Latest Block change to "a few seconds ago"?

What do you get when you hit the back button? What is your report below the Node On button?

Also is your device on 24/7?

Hey. Thanks for the response.

Yes, for synced and latest block change a few seconds ago.

Yes, for running 24/7

When clicking the back button the node shows as being turned on, running the blockchain, with both the latest block number and the local block number.

Everything appears to be running as should be, except there has been no incoming connections detected, still. I am not clear as to whether that is usual or problematic, as I am absolutely not an expert, either.

[steeveee100]

Yes could be usual, not problematic. You're actually still very new, just been up a few days, usual guidance is wait! Your availability is good - I've been up nearly a month, at first my availability shot up to 60 plus %, now barely gains .5% a day.

Just to check your ports again. Turn off the Node using the switch, go back to the Port Checker and run it. Does it report all ports open?

[anon-netizen]

Yes could be usual, not problematic. You're actually still very new, just been up a few days, usual guidance is wait! Your availability is good - I've been up nearly a month, at first my availability shot up to 60 plus %, now barely gains .5% a day.

Just to check your ports again. Turn off the Node using the switch, go back to the Port Checker and run it. Does it report all ports open?

Ah, yes, thanks. Ugh, I had not thought to do this! A problem was discovered. The ping back being unsuccessful. However, when I check the firewall, the designated ports all appear to be open.

[anon-netizen]

No doubt I am overlooking something, somewhere.

[steeveee100]

Ok, what happens when you run https://www.ipfingerprints.com/portscan.php? All ports should be marked as open.

[nathusius]

@anon-netizen You might need to restart the "stellar-dummy" container from within Docker before the built-in Port Checker function works, but also see advice above: the built-in portscan check is temperamental. You can sanity-check your ports at https://www.ipfingerprints.com/portscan.php (you do need to manually start stellar-dummy for this to work, I just tried it myself!) For me the built-in port checker never even attempted to test anything, but ipfingerprints worked fine with all ten ports showing as open and I ended up getting picked up OK. Did you need to setup port forwarding on your router? If so, re-check that this is setup correctly, as I found mine got messed up after a reboot and I needed to allocate a fixed IP address to my network adapter, so that the port forwarding as still valid after a PC reboot (this is for wired connection to a router)

[anon-netizen]

@anon-netizen You might need to restart the "stellar-dummy" container from within Docker before the built-in Port Checker function works, but also see advice above: the built-in portscan check is temperamental. You can sanity-check your ports at https://www.ipfingerprints.com/portscan.php (you do need to manually start stellar-dummy for this to work, I just tried it myself!) For me the built-in port checker never even attempted to test anything, but ipfingerprints worked fine with all ten ports showing as open and I ended up getting picked up OK. Did you need to setup port forwarding on your router? If so, re-check that this is setup correctly, as I found mine got messed up after a reboot and I needed to allocate a fixed IP address to my network adapter, so that the port forwarding as still valid after a PC reboot (this is for wired connection to a router)

Thank you very much for this response! I will methodically work through all these advices; see what the outcome is and post my findings later on, today.

[anon-netizen]

re: No I have not specifically reconfigured my router, as yet. Only going to the extent of opening the dedicated ports in my Firewall app.

[anon-netizen]

Ok, what happens when you run https://www.ipfingerprints.com/portscan.php? All ports should be marked as open.

Thanks, I will do this, find out & report back. Presently I am now on a work call and have to pause my investigation for the moment.

[nathusius]

re: No I have not specifically reconfigured my router, as yet. Only going to the extent of opening the dedicated ports in my Firewall app.

Try ipfingerprints first, if the ports are not all showing as open (it will show closed or filtered instead), then you will probably need to investigate port forwarding on the router!

[ihatejam]

Ok. I'm not an expert, but does State ever say "Synced!"? And Latest Block change to "a few seconds ago"? What do you get when you hit the back button? What is your report below the Node On button? Also is your device on 24/7?

Hey. Thanks for the response.

Yes, for synced and latest block change a few seconds ago.

Yes, for running 24/7

When clicking the back button the node shows as being turned on, running the blockchain, with both the latest block number and the local block number.

Everything appears to be running as should be, except there has been no incoming connections detected, still. I am not clear as to whether that is usual or problematic, as I am absolutely not an expert, either.

With what you've reported with being sync'd to the latest block -- Everything is perfect! No need to do anything else.

[b2344]

[b2344]

It's one day stay in the picture ，my local blockchain never move

[ihatejam]

Click the blue Troubleshooting link. On the Troubleshooting screen, verify that you are running version 0.44. After that, click the "Remove blockchain data" link and then give it an hour or two.

[pawelwegner]

Can you share the data.. troubleshooting + results from command docker ps -a and wsl -l -v from command prompt?

All seems OK, but I'm worried about N/A in the availability stats...

Can anyone can tell me where to find the error log? Where can I find out what the consensus container error is? (it changes error <-> absent all the time)

[nathusius]

@pawelwegner Have a look at some other recent threads - I don't think that anyone has got the Node to work sucessfully with Windows Server!!

[ihatejam]

@pawelwegner there's a long history of issues with running Linux docker containers on Win Server. As @nathusius mentioned - don't think anyone has the node running on Win server. Also, as I've mentioned before in other threads -- you don't need a server class machine -- just a desktop with a recent build of win 10 with docker desktop should be fine.

[If you still have issues, you may want to start a new thread]

[github-cloudlabsuser-078]

Running a Windows Server node: Availability (up to 90 days): 67.74% from 2021-04-05 to 2021-05-17 (not so bad).

Main issue: PI app not managed as a Windows service (server automated restart not managed) Other technical issue: related to docker on Windows Server constraint on Hyper-V way of managing storage. Resolved: Using a scheduled task to restart the node each time a docker event 11 is raised.

[hawjiang5349]

Availability (up to 90 days): 69.20% from 2021-04-21 to 2021-07-20 Consensus container: Creating/absent

Same problem here after update latest docker.

[hawjiang5349]

Problem still exist since 21 days ago updated latest docker. 3 weeks and 24/7 already. I will keep updating the issue here. Node installation date: 2021-04-08 Availability (up to 90 days): 68.30% from 2021-04-28 to 2021-07-27 Consensus container: Creating/absent

[ihatejam]

@hawjiang5349 It might be better to start a separate Github thread with your issue in it and provide your own set of data -- Is your node reachable from outside of your home network -- Are the ports forwarded properly? What does https://www.ipfingerprints.com/portscan.php report for the ports that are meant to be forwarded? (Please do not post pictures containing your public/external IP addresses) Do you leave your Node running all the time? If your node is unreachable, then it will not get the consensus container.