-
Notifications
You must be signed in to change notification settings - Fork 20k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
leveldb/table: corruption on data-block #2568
Comments
I have the same problem, but here it seems to be related to the ARMv7hf architecture.
|
It is not limited to arm and windows. I am trying to sync on Ubuntu Linux server 15.04 with geth --fast console 3 days and always get this error on different blocks. On the other hand on other machine geth sync as it should. I tried to change hard disk, bought fresh new SSD and error is here again. I don't know what to do.I have celeron 2 core CPU and 4GB of RAM. I have 128 GB SSD. CPU and OS: After that error I can only close hang ssh connection, when I press key nothing happens. Just ctrl+c give me new prompt on which I can't type anything... Last parts of log in chaindata after another attempt after removing chain data are: 08:57:40.185549 table@compaction L1·1 -> L2·6 S·14MiB Q·10503753 |
I'm also experiencing this issue, running on a rPi 2 - chaindata writing to a USB mount. Geth version 1.3.3 F0630 22:54:51.446400 3865 database_util.go:362] failed to store block body into database: leveldb/table: corruption on data-block (pos=2028406): checksum mismatch, want=0x4b0adb2e got=0x38d5ebdd [file=015275.ldb] |
This error is still present in version: 1.4.9-stable-b7e3dfc5 |
I also got the error (Version 1.4.10 and 1.4.11). |
Would like to report experiencing this issue too. Seems to happen after starting/stopping My
|
We have moved to a newer Version of leveldb in geth 1.5.0. |
I'm having the "same" issue with geth 1.5.7 when I run Info
Details
|
Same problem with last stable (making sync or import will produce error):
|
Getting similar problem with gets 1.6.6 on Mac This is after I deleted the blockchain because node could not sync for another problem... elapsed=236.740ms processed=9723275 pending=601 retry=0 duplicate=47879 unexpected=132572 Any ideas? (Checked SSD smart status - no relocated sectors added...) |
Windows 8.1 having the same issue using --fast and also when not. Cannot sync. Cannot use Ethereum. Geth/v1.6.6-stable-10a45cb5/windows-amd64/go1.8.3 CRIT [07-02|20:58:35] Failed to store block body err="leveldb/table: corruption on data-block (pos=1020486): checksum mismatch, want=0xd897cf55 got=0xa9b3f8d2 [file=035242.ldb]" |
same issue CRIT [07-12|00:11:22|core/database_util.go:329] Failed to store last fast block's hash err="leveldb/table: corruption on data-block (pos=759124): checksum mismatch, want=0x7fe9fb78 got=0x29fa1df7 [file=126096.ldb]" |
Same issue appears for me,
Built from source on MacOS Sierra at commit:
Is there any other state I could collect that would aid the developers in debugging? |
@gorilla2112 Any motivation about close this issue ? |
why is this issue closed? am i missing something? or doing something terribly wrong? Still getting this problem with tried deleting blockchain multiple times...
|
In my case, I deleted the corrupt file (such as 102005.ldb) and restarted geth --fast. It then continued syncing. I am not sure if this is how it should be, I can just hope that the leveldb files are sort of stream like and it will resync everything that was in the deleted file (102005.ldb) on start. Any confirmation from anyone that this is a good solution? |
Still getting this issue with geth 1.7.3 on Ubuntu 17.10 (on HDD) and on the same machine with macOS High Sierra (on SSD), and Windows 10 (on a different SSD) using |
I found my problem to be the power supply. I tried different memory sticks first which had no effect, and then unplugged the extra hard drives and it worked better. I was pulling too much power from my 600w supply. Recently added an RX 580 and another SSD. Upgraded to an 850w and it is syncing much faster now without issues. |
Getting the same error on Ubuntu 14.04, Geth – Weir (v1.7.3): Failed to store last header's hash err="leveldb/table: Writer: keys are not in increasing order |
I was able to solve the problem on windows after 5 days trying to synch.
*Synch took about 5 hours on a new ssd / 100Mbit connection Not sure which specific action made me successfully synch, I have done all the above after many days of trying to synch the damn node, I even bought new SSD and it did not help. Hope you find my solution to work for you lost souls. btc: 12y5cQN8RcmUgatn71UbD6eMWhHGCHu8fE |
have the same issue geth 1.8.0
seriously it is 5th time I am trying to sync geth and no luck. another issue again. geth is not syncing. just deleting the file leads to this issue: |
There are some comments earlier in the thread that suggest this error might come up for all classes of memory corruption. Having a bad memory module or a noisy/overloaded power supply, or a weakly-connected SATA cable might all be the culprit -- but it's probably some kind of extremely rare intermittent hardware failure, even on a machine that otherwise "seems okay". Flipping one bit in a billion won't impair an MPEG movie, but it will make a checksum fail. |
there are no cables. ssd is integrated into motherboard. the power supply uses a battery. |
@winsvega did you get a solution for this? I have the same problem .. tried resetting the complete setup for more than 3 times now, but landing on the same issue again n again ... Looks like a very common problem, but no where I can find a clean solution .. :( Not sure why this ticket is closed! |
I heard the go team has fixed the light client again. so try syncing with the geth light client option enabled. my solution was to switch to the parity client. as they are focusing their efford on faster blockchain syncing. |
Is there a workaround / solution for this bug? I get this error when attempting to fast sync geth on mainnet:
Edited to add -- I'm running the latest master version of geth on Xubuntu 16.04:
|
no. either light mode or full resync. |
I did a full resync, and everything worked this time... very mysterious. |
I have the same issue, deleting the corrupted file only works for a bit before the next file is corrupted. This issue is now two years old and it doesn't seem like this will be fixed anytime soon. Why can't the node just delete the file and try a resync from there on instead of crashing? |
this problem is definitely still present. |
get the same with 1.8.27 stable and 1.9 unstable, e.g.
|
the real bug here is the error message saying there is a bug.
this is a subtle hardware problem and can be the motherboard, CPU, memory
or SSD drive. do try on different hardware, worked for me.
…On Wed, May 22, 2019, 4:30 AM Alex Ivasyuv ***@***.***> wrote:
get the same with 1.8.27 stable and 1.9 unstable, e.g.
CRIT [05-21|21:09:21.895] Failed to store hash to number mapping
err="leveldb/table: corruption on data-block (pos=91219): checksum
mismatch, want=0x52086fd8 got=0x17a37886 [file=490521.ldb]"
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2568?email_source=notifications&email_token=AAFJBNDPMNSBDN7NGZ43VMDPWR5IVA5CNFSM4CD2BUO2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODV5NMMY#issuecomment-494589491>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAFJBNCTLWCLFW6YPGX4LIDPWR5IVANCNFSM4CD2BUOQ>
.
|
didn't point, that it's mainnet. running on official docker. current size of chain ~ 130GB synced ~88%. |
I got this error today after downloading blockheaders for 2 weeks , I have 1 T SSD, occupied nearly 400GB now. :-( ran geth prior to latest Geth/v1.9.24-stable-cc05b050/windows-amd64/go1.15.5 got the error any help pls |
Not sure if this helps any one, You need to delete the data at the cache dir. Somewhere in the sync the program broke, or had a partial download. Clearing your data dir which you can find with |
A simpler way is to do |
It is likely that the memory is corrupted if the disk SMART reports no corruption. Here are my check steps (on Linux):
(Different disk may report different formats, and mine is 970 evo pro). The report tells that no integrity errors are found, so it is unlikely the corruption issue from disk.
Further, if you want more reliable memory, you could always choose ECC ones. |
System information
Geth version:
0.9.41
OS & Version: Windows10
Expected behavior
connection to blockchain
Actual behavior
Error:
eth/downloader/downloader.go:274] Synchronisation failed: leveldb/table: corruption on data-block (pos=0): checksum mismatch, want=0x8f40a97e got=0x792b89e6 [file=4762443.ldb]
Steps to reproduce the behavior
geth --rpc
The text was updated successfully, but these errors were encountered: