Jun 20 12:56:56 xxxxxxxxxx kernel: blk_insert_cloned_request: over max segments limit. (272 > 256)
Jun 20 12:56:56 xxxxxxxxxx kernel: VxVM vxdmp V-5-3-0 dmp_send_scsipkt_cloned_req: failed to insert cloned request
A FS thread is flushing pages in a shared state is stuck flushing. This is due to changes made to provide support for BLK_MQ for RHEL9. VxVM is sending requests which cannot be honored by the IO layers below.
crash> bt ff2b505d075e2340
PID: 410 TASK: ff2b505d075e2340 CPU: 19 COMMAND: "kworker/u130:3"
#0 [ff6389f7cd497680] __schedule at ffffffff9586658a
#1 [ff6389f7cd4976e8] schedule at ffffffff95866a2d
#2 [ff6389f7cd4976f8] vx_io_startnowait at ffffffffc16cbad4 [vxfs]
#3 [ff6389f7cd497778] vx_flush_pages_lnx at ffffffffc16cc156 [vxfs]
#4 [ff6389f7cd497878] vx_pvn_range_dirty at ffffffffc16cd05c [vxfs]
#5 [ff6389f7cd4979b8] vx_putpage_dirty_wbc at ffffffffc16ce759 [vxfs]
#6 [ff6389f7cd497a18] vx_do_putpage_wbc at ffffffffc16ce9a3 [vxfs]
#7 [ff6389f7cd497a40] vx_putpage1 at ffffffffc1698c2d [vxfs]
#8 [ff6389f7cd497a98] vx_writepages at ffffffffc166731d [vxfs]
#9 [ff6389f7cd497b28] do_writepages at ffffffff94f2ceac
#10 [ff6389f7cd497bc0] __writeback_single_inode at ffffffff9505f3d1
#11 [ff6389f7cd497c00] writeback_sb_inodes at ffffffff9505fae9
#12 [ff6389f7cd497cd8] __writeback_inodes_wb at ffffffff9505fddc
#13 [ff6389f7cd497d18] wb_writeback at ffffffff95060057
#14 [ff6389f7cd497d88] wb_do_writeback at ffffffff9506038a
#15 [ff6389f7cd497e00] wb_workfn at ffffffff9506137e
#16 [ff6389f7cd497e90] process_one_work at ffffffff94d2b892
#17 [ff6389f7cd497ed8] worker_thread at ffffffff94d2be60
#18 [ff6389f7cd497f18] kthread at ffffffff94d33d0d
#19 [ff6389f7cd497f50] ret_from_fork at ffffffff94c03129
Disable BLK_MQ support form DMP.
sysctl -w vxdmp.dmp_blk_mq_enable=0 >> /etc/sysctl.conf
reboot
Veritas is currently working on a permanent fix. Contact Veritas support for the current status.