Node losing sync and needs manual connect to peers [BUG] #3298
Comments
|
Have the same issue. It started to happen yesterday. I also have two plotting PCs, different wallets. One is syncing good, the second keep losing peers. I can see that its trying to connect but failing. |
|
same |
|
same here |
|
Same |
|
same here |
|
same. i've not been able to sync since 4-27. |
|
Yea the syncing is pretty bad. Sometimes I randomly check my farmer and it is not sycned and has no peers. I try to copy paste node IPs and port from my other farmer and nothing happens... pretty lame. |
|
same. lose one night while waiting sync. |
|
1.1.3 was released, and at least it solved my sync problems |
|
1.1.3 solved the problem, do not use 1.1.4. I confirm this. |
|
Getting the same. |
|
Had the same problem on 1.1.3. Updated to 1.1.4 and still having the same problem. |
|
Having the same problem. In Germany it is common being disconnected by the ISP after 24hr. That said, after disconnect I'm facing the resync issue. I have to close the Chia App and start it again. |
|
Same. |
|
fully portmapped system that was set with introducer-or.chia.net (my closest as per #support macrobot on keybase) lost sync today with over 52 peers. checking and all of them were below my height. this should not be possible unless i am the only top node, which is obviously not the case. this doesn't bode well for the longevity of xch unless they can more smartly design their node ranking and introducer system. |
|
Same here. Using 1.1.4 Suddenly there is a Full Node connecting, Now it's syncing.. But for how long? |
|
same
|
|
Same issue here. So, is this ever going to get fixed? |
|
I'm still on 1.1.3 and its synced for 20 mins then not for 20 mins. Driving me mad as I have to time with plotting to update and restart but not farming while messing about. |
|
Agree. Same problem. Just started a day ago. Checked ports, 8444 open. Trying to reset, uninstall etc etc. done it all. Very frustrating. |
|
Just for those here, my workaround for the problem of plotting and farming on the same device, is to personally prioritize plotting now. When I see a de-sync, I use this site (https://chia.keva.app/) to find new IP addresses, and run through all the It's manual, and a pain, and doesn't always work or stay stable, BUT it prevents me from having to restart plot building. |
|
Please do something about sync - there is absolutely no way i can forward port 8444, my ISP prevents that. Please make outgoing connections better atleast! |
|
Same here. Once a day the fullnode and the wallet go out of sync and even after waiting for 3h the full node nor the wallet got synced again. I am running 1.1.15 |
|
Yup, same here. Made it work adding random introducers, but this will be hurtfull for chia when average users will kick in. |
|
Well, I'm about done with this BS. I finally figured out a work around to open port 8444 on my router and it seemed that it had mitigated my constant sync\not available issues. Nope. Worked for about 4 hours, and then all 243 plots went unavailable again. So I tried adding the suggested peer connections (again) to try and re-sync, but it froze up Chia (endless spinning circle for an hour) losing 10 plots that were over 50%, and causing me to nearly start chucking 12tb drives out the friggin' window. I, for the life of me, do not understand why farming and plotting have to be tied at the hip. I even went as far as to load Ubuntu this morning, but I couldn't even get Chia to even want to try to sync. It was worse. Really no point in wasting money\time on plots if you can't farm the damn things. It'd be different if this was an isolated issue, but it appears that a significant portion of users are experiencing the same problems. Please fix this on your next update. |
|
I agree, that's a major issue. |
|
I'm pretty frustrated with this issue also. There should be a file where we can put all those working nodes and have the Chia client runs thru the list on its own, instead of manually copy n paste all the IPs, and some of them don't ever work!!!!! |
|
What might be a work around if someone is ambitious, would be to build a site like the one built for Pokemon Go that lets people post friends codes to. It might be better than https://chia.keva.app/ . https://pokemongo.gishan.net/friends/codes/ Really though, this is a pretty core problem and needs to be addressed directly by the developers. I should be able to set a default connection number, and have the application manage the connections to keep that many connections up. I personally wonder exactly what the Total Network Space would be if node didn't drop off so regularly. |
|
Same issue with me, after about 2 days of waiting I added the node-eu.chia.net:8444 to peers again and after about 30minutes the Sync started moving again. Its now finished |
|
Tried many things in settings and agree with loads of comments. In the end, I found a cheap second hand pc with a better processor (from i3 to i7) actually less RAM but similar in every other way. The better processor managed to sync albeit slowly. |
|
I have a Raspberry Pi 4 as a separate farmer (no plotting made on it, I
have a separate harvester) and the issue is still here on the farmer.
Stefano Scarpone
Il mer 19 mag 2021, 13:33 gregoriob ***@***.***> ha scritto:
… Tried many things in settings and agree with loads of comments. In the
end, I found a cheap second hand pc with a better processor (from i3 to i7)
actually less RAM but similar in every other way. The better processor
managed to sync albeit slowly.
Other potential was that the i3 laptop was confusing itself with WiFi &
Ethernet trying to run in parcelled.
If you read guides on farming with multiple PCs, it states that splitting
farming & plotting separately is a better use of resource. Having one
farmer with the blockchain synced up managing the farm & secondary (or
more) plotting might be the work around.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3298 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AA2UKYPVWWTG3UTRHIFNYX3TOOOW3ANCNFSM436C5RVQ>
.
|
You don't need your farmer to sync from what I understand. You should only need one db in your internal network. Duplicating causes issues, just disable upnp on the farming node? |
|
My Raspberry Pi is both the farmer and full node, while I have a separate
PC as an harvester (which is then not syncing with the blockchain but only
provide the plots), I disabled UPNP and I have a manual port forwarding
towards port 8444 of the Raspberry, indeed I can see nodes connecting to me
correctly and the blockchain height is increasing but I never reach the top
of the chain anyway. This was not happening since few days ago, so it's
very weird.
Il giorno mer 19 mag 2021 alle ore 14:11 gregoriob ***@***.***>
ha scritto:
… I have a Raspberry Pi 4 as a separate farmer (no plotting made on it, I
have a separate harvester) and the issue is still here on the farmer.
Stefano Scarpone
Il mer 19 mag 2021, 13:33 gregoriob *@*.***> ha scritto:
Tried many things in settings and agree with loads of comments. In the
end, I found a cheap second hand pc with a better processor (from i3 to i7)
actually less RAM but similar in every other way. The better processor
managed to sync albeit slowly.
Other potential was that the i3 laptop was confusing itself with WiFi &
Ethernet trying to run in parcelled.
If you read guides on farming with multiple PCs, it states that splitting
farming & plotting separately is a better use of resource. Having one
farmer with the blockchain synced up managing the farm & secondary (or
more) plotting might be the work around.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#3298 (comment)
<#3298 (comment)>
,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AA2UKYPVWWTG3UTRHIFNYX3TOOOW3ANCNFSM436C5RVQ
.
You don't need your farmer to sync from what I understand. You should only
need one db in your internal network. Duplicating causes issues, just
disable upnp on the farming node?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3298 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AA2UKYMWBDR7LZJMWWHQ5KTTOOTG7ANCNFSM436C5RVQ>
.
--
Stefano Scarpone
|
Same here. Exactly the same problems. |
|
I actually brought this up a month ago on Keybase. Exact same problem. I thought it was the other nodes. |
|
+1 |
|
I deleted the peer_table_node.sqlite and turned off disk sleep and now it works for me. |
|
Deleting .chia/mainnet/db/peer_table_node.sqlite worked for me as well. This forces the client to request new peers. |
|
1.1.6 claims to resolve this now... let's see |
|
It's off to a strong start! |
|
Ja, right after the full-node start I got 7 connections to other full nodes. Catching up to the newest block is still a little bit slow in my opinion ... My previous idea, that the disconnect has something to do with a internal full queue of bad peers was wrong. I am hoping that this release will fix the situation fingers crossed |
|
Upgrading to 1.1.6 (and deleting .chia/mainnet/db/peer_table_node.sqlite
too to be sure) finally solved the issue for me!
Il giorno ven 21 mag 2021 alle ore 09:07 Markus Mazurczak <
***@***.***> ha scritto:
… Ja, right after the full-node start I got 7 connections to other full
nodes. Catching up to the newest block is still a little bit slow in my
opinion ...
My previous idea, that the disconnect has something to do with a internal
full queue of bad peers was wrong.
My full nodes disconnects every night between 3:12 am and 3:27 am (which
sounds a bit weird) but at the point when it starts trying to connect to
only one peer for hours the internal queue was empty and even there were no
blocked bad peers.
I am hoping that this release will fix the situation *fingers crossed*
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3298 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AA2UKYMHYTUOXTLLOUYUDYDTOYBCRANCNFSM436C5RVQ>
.
--
Stefano Scarpone
|
Confirm, i think solved, but i look next 24 hours for sure... |
Not likely... I've observed stablized sync since last Fri 5/14, with no changes client-side across 12 systems at 4 different geographic locations. More likely, the exponential ramp-up of new zero-height clients signing-on/synching and dragging down all the other clients on the net has subsided. It's too bad no one took this issue seriously when it was occurring, else we might have been able to figure out a solution for when the net is faced with a similar issue. Maybe it will face that issue again... but we may never know. Let's hope not. With such a large number of clients at full-height now, it takes a much larger-and-larger numbers of zero-height clients connecting/min to hit that critical point, so hopefully those days are behind us. Still, it's worrisome in case a bad actor might be able to exploit the vulnerability, but given the architecture, I think we are safe. One would likely have resources to execute a 51% attack that would be much more useful than simply causing some <<50% of nodes to go out of sync. |
|
Seems like the problem with that disconnect is solved. I have not have any issues for about 3 days now. No disconnect |
I am too, its now working like normally. |
|
same. stable few days on 1.1.6
in fact the most connections since plots completed.
sadly the horses are out of the barn, smaller farming was just eclipsed
during this whole ordeal.
…On Sun, May 23, 2021 at 11:16 AM Andrey Kuvshinov ***@***.***> wrote:
Seems like the problem with that disconnect is solved. I have not have any
issues for about 3 days now. No disconnect
Im too, its now working like normally.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3298 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AT5GE4UMPEUHXY44PLWJNIDTPES47ANCNFSM436C5RVQ>
.
|
Deleting this file helped me as well. Thank you the post. |
|
Same here
…On Sun, May 23, 2021, 9:16 AM Andrey Kuvshinov ***@***.***> wrote:
Seems like the problem with that disconnect is solved. I have not have any
issues for about 3 days now. No disconnect
Im too, its now working like normally.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#3298 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AUAGMNGF2OB2WP47VKCVK4DTPES47ANCNFSM436C5RVQ>
.
|
|
deleted this file, wallet still doesn't sync reliably at all |
|
in the time it took my node to sync from 0 to full, my wallet is at a block height of 25888 with no sign of budging. |
|
Sometimes your peer list will fill with nodes that are at a lower block height. Because of this, there's nobody that you can sync from, only sync to. The only solution is to restart your full node (which is taking an ever-increasing amount of time due to chain growth) and hope you get better nodes next time. |
|
This issue has been flagged as stale as there has been no activity on it in 14 days. If this issue is still affecting you and in need of review, please update it to keep it open. |
github-actions
bot
added
the
stale-issue
|
This issue was automatically closed because it has been flagged as stale and subsequently passed 7 days with no further activity. |
and others
Describe the bug
In the GUI the node loses sync several times a day. It can only be reconnected by manual connection to peers and doesn't look up peers on its own. If a manual connect is made, it occasionally detects another peer automatically.
To Reproduce
Leave the GUI alone plotting...
When it loses sync, try to reconnect.
Expected behavior
It is expected that the node can keep sync itself
Screenshots
Desktop (please complete the following information):
Additional context
Luckily I have two plotting computers. The other computer remains synced by itself and has a list of peers.
Usually, both node.chia.net and chiadb.net are occupied when I try to connect, but I can find another 8444/8444 IP-address in the list of my second computer to use for the unsynced computer. Unfortunately, it usually takes a long time to find a node that accepts connections...
This bug is a deathstroke to Chia. If people have a hard time even winning blocks, then it won't be easier if they cannot sync...
The text was updated successfully, but these errors were encountered: