PXC 5.7 xtrabackup: Error writing file 'UNOPENED' (Errcode: 32 - Broken pipe)

Hello Everyone,

I have an issue described by the sentence above. What happened is that we managed to break one node of a 2 node cluster, so I put node one back to bootstrap, preinstalled node2 but now it can’t sync as xtrabackup on the bootstrapped (donor) node gives the message from the subject on completely random places.

The sync worked perfectly the first time, but now that I preinstalled node2, mysql starts, starts to receive the data from node1 and and then this message comes up every time, but with random tables. I’ve checked several tables on the donor side with mysqlcheck and they seem to be allright.

Any hints?

BTW, it’s a ~600Gb size database.

Here are parts of the log donor side:

180803 03:38:10 >> log scanned up to (1887066912173)
180803 03:38:11 >> log scanned up to (1887066912173)
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
xb_stream_write_data() failed.
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
[04] xtrabackup: Error: xtrabackup_copy_datafile() failed.
[04] xtrabackup: Error: failed to copy datafile.
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
xb_stream_write_data() failed.
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
[02] xtrabackup: Error: xtrabackup_copy_datafile() failed.
[02] xtrabackup: Error: failed to copy datafile.
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)*** Error in `xtrabackup’: corrupted size vs. prev_size: 0x00000000023e8c00 ***

xb_stream_write_data() failed.
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
[03] xtrabackup: Error: xtrabackup_copy_datafile() failed.
[03] xtrabackup: Error: failed to copy datafile.
xtrabackup: ======= Backtrace: =========
Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
xb_stream_write_data() failed.
/lib64/libc.so.6(+0x7f5e4)[0x7f8c6f6a05e4]
/lib64/libc.so.6(+0x816db)[0x7f8c6f6a26db]
xtrabackup: Error writing file ‘UNOPENED’ (Errcode: 32 - Broken pipe)
/lib64/libc.so.6(+0x39c00)[0x7f8c6f65ac00]
[01] xtrabackup: Error: xtrabackup_copy_datafile() failed.
[01] xtrabackup: Error: failed to copy datafile.
/lib64/libc.so.6(+0x39c27)[0x7f8c6f65ac27]
.
.
.
.
7ffcc637d000-7ffcc637f000 r-xp 00000000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall]
08:38:12 UTC - xtrabackup got signal 6 ;
This could be because you hit a bug or data is corrupted.
This error can also be caused by malfunctioning hardware.
Attempting to collect some information that could help diagnose the problem.
As this is a crash and something is definitely wrong, the information
collection process might fail.

Thread pointer: 0x0
Attempting backtrace. You can use the following information to find out
where mysqld died. If you see no messages after this, something went
terribly wrong…
stack_bottom = 0 thread_stack 0x10000
xtrabackup(my_print_stacktrace+0x2c)[0xe627fc]
xtrabackup(handle_fatal_signal+0x262)[0x9e27e2]
/lib64/libpthread.so.0(+0xf6d0)[0x7f8c719dc6d0]
/lib64/libc.so.6(gsignal+0x37)[0x7f8c6f657277]
/lib64/libc.so.6(abort+0x148)[0x7f8c6f658968]
/lib64/libc.so.6(+0x78d37)[0x7f8c6f699d37]
/lib64/libc.so.6(+0x7f5e4)[0x7f8c6f6a05e4]
/lib64/libc.so.6(+0x816db)[0x7f8c6f6a26db]
/lib64/libc.so.6(+0x39c00)[0x7f8c6f65ac00]
/lib64/libc.so.6(+0x39c27)[0x7f8c6f65ac27]
xtrabackup[0x72ae97]
/lib64/libpthread.so.0(+0x7e25)[0x7f8c719d4e25]
/lib64/libc.so.6(clone+0x6d)[0x7f8c6f71fbad]

Please report a bug at [URL]Log in - Percona JIRA
--------------- innobackup.backup.log (END) ----------------------

Hello.

Did you find a fix for this issue?

I am experiencing the same one (o very similar).

I have reported it here:

[url][PXB-1677] Cannot add a new node to a cluster because SST fails - Percona JIRA

Hello ofrias, I just checked in with the Percona XtraBackup team, and they are taking a look at the JIRA.

Could you update the JIRA [URL][PXB-1677] Cannot add a new node to a cluster because SST fails - Percona JIRA with any error logs from your server, network logs whatever, since they need to know what caused the broken pipe. It could be a network error for example.

Thanks for taking the time to report this by the way.

They closed it as a duplicate. My story with it ended with droping this solution