0

I have a problem with Percona 8 on Debian 11, I installed the latest version of Percona 8 for MySQL server, 8.0.29-21

I set these configurations in the mysqld.cnf

innodb_buffer_pool_size = 12G innodb_log_buffer_size= 128M sql-mode="" join_buffer_size = 4M sort_buffer_size = 2M max_connections = 800 max_allowed_packet= 128M thread_handling=pool-of-threads innodb-rollback-on-timeout = ON net_read_timeout = 60 net_write_timeout = 60

I have a backup of my database with size 11GB compressed, I tried to restore the backup but I keep get the following Error

ERROR 2013 (HY000) at line 3867: Lost connection to MySQL server during query

after changing some variables like read and write timeouts, I did not work, then I noticed that the Percona server service is restarted by itself while restoring the data

and I have another server also running on Proxmox VM with same settings and the restore of the database is working fine

and I found these error in the syslog file

#####################

Nov 13 02:43:57 rms-new kernel: [262131.431502] sd 0:0:1:0: [sdb] tag#9 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE cmd_age=0s
Nov 13 02:43:57 rms-new kernel: [262131.431523] sd 0:0:1:0: [sdb] tag#9 Sense Key : Illegal Request [current] 
Nov 13 02:43:57 rms-new kernel: [262131.431526] sd 0:0:1:0: [sdb] tag#9 Add. Sense: Invalid field in cdb
Nov 13 02:43:57 rms-new kernel: [262131.431528] sd 0:0:1:0: [sdb] tag#9 CDB: Write(10) 2a 00 08 67 14 00 00 0a 00 00
Nov 13 02:43:57 rms-new kernel: [262131.431532] blk_update_request: critical target error, dev sdb, sector 140973056 op 0x1:(WRITE) flags 0x4800 phys_seg 20 prio class 0
Nov 13 02:43:57 rms-new kernel: [262131.433123] EXT4-fs warning (device sdb): ext4_end_bio:347: I/O error 5 writing to inode 19136514 starting block 17622016)
Nov 13 02:43:57 rms-new kernel: [262131.433182] buffer_io_error: 1014 callbacks suppressed
Nov 13 02:43:57 rms-new kernel: [262131.433184] Buffer I/O error on device sdb, logical block 17620992
Nov 13 02:43:57 rms-new kernel: [262131.433496] Buffer I/O error on device sdb, logical block 17620993
Nov 13 02:43:57 rms-new kernel: [262131.433711] Buffer I/O error on device sdb, logical block 17620994
Nov 13 02:43:57 rms-new kernel: [262131.433921] Buffer I/O error on device sdb, logical block 17620995
Nov 13 02:43:57 rms-new kernel: [262131.434132] Buffer I/O error on device sdb, logical block 17620996
Nov 13 02:43:57 rms-new kernel: [262131.434356] Buffer I/O error on device sdb, logical block 17620997
Nov 13 02:43:57 rms-new kernel: [262131.434562] Buffer I/O error on device sdb, logical block 17620998
Nov 13 02:43:57 rms-new kernel: [262131.434780] Buffer I/O error on device sdb, logical block 17620999
Nov 13 02:43:57 rms-new kernel: [262131.434988] Buffer I/O error on device sdb, logical block 17621000
Nov 13 02:43:57 rms-new kernel: [262131.435197] Buffer I/O error on device sdb, logical block 17621001
Nov 13 02:43:57 rms-new kernel: [262131.458839] kauditd_printk_skb: 6 callbacks suppressed
Nov 13 02:43:57 rms-new kernel: [262131.458841] audit: type=1400 audit(1668325437.527:1437): apparmor="ALLOWED" operation="open" profile="/usr/sbin/mysqld" name="/proc/3310/task/3419/mem" pid=3310 comm="connection" requested_mask="r" denied_mask="r" fsuid=106 ouid=106
Nov 13 02:43:57 rms-new systemd[1]: mysql.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
Nov 13 02:43:57 rms-new systemd[1]: mysql.service: Failed with result 'exit-code'.
Nov 13 02:43:57 rms-new systemd[1]: mysql.service: Consumed 5min 54.281s CPU time.
Nov 13 02:43:57 rms-new systemd[1]: mysql.service: Scheduled restart job, restart counter is at 1.
Nov 13 02:43:57 rms-new systemd[1]: Stopped Percona Server.
Nov 13 02:43:57 rms-new systemd[1]: mysql.service: Consumed 5min 54.281s CPU time.
########################

1 Answer 1

1

I found the problem,

it was the disks added by the VM provider, they are using hardware not compatible with new SSD disks, when they moved the VM to the new node compatible with the storage, problem solved

Thanks

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.