Multipath Failure with Error "blk_cloned_rq_check_limits: over max size limit" (Doc ID 2524139.1)

Last updated on JULY 02, 2021

Applies to:

Symptoms

 A system running  Oracle Linux 7 and using EMC XTREM I/O storage experiences multipath failures when booted into kernel 4.1.12-94.5.9.el7uek.x86_64.

The following multipath messages are seen in the system logs:
Jan 2 17:47:34 server multipathd: mpathah: remaining active paths: 3
Jan 2 17:47:34 server multipathd: mpathah: remaining active paths: 2
Jan 2 17:47:34 server multipathd: mpathah: remaining active paths: 1
Jan 2 17:47:34 server multipathd: mpathah: remaining active paths: 0
Jan 2 17:47:34 server multipathd: mpathw: remaining active paths: 3
Jan 2 17:47:34 server multipathd: mpathw: remaining active paths: 2
Jan 2 17:47:34 server multipathd: mpathw: remaining active paths: 1
Jan 2 17:47:34 server multipathd: mpathw: remaining active paths: 0
Jan 2 17:47:34 server multipathd: mpathas: remaining active paths: 3
Jan 2 17:47:34 server multipathd: mpathas: remaining active paths: 2
Jan 2 17:47:34 server multipathd: mpathas: remaining active paths: 1
Jan 2 17:47:34 server multipathd: mpathas: remaining active paths: 0
Jan 2 17:47:34 server multipathd: mpathah: sdid - tur checker reports path is up
Jan 2 17:47:34 server multipathd: mpathah: remaining active paths: 1
Jan 2 17:47:38 server multipathd: mpathw: sdj - tur checker reports path is up
Jan 2 17:47:38 server multipathd: mpathw: remaining active paths: 1
Jan 2 17:47:38 server multipathd: mpathah: sdm - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathah: remaining active paths: 2
Jan 2 17:47:38 server multipathd: mpathas: sdp - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathas: remaining active paths: 1
Jan 2 17:47:38 server multipathd: mpathw: sdcg - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathw: remaining active paths: 2
Jan 2 17:47:38 server multipathd: mpathah: sdcj - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathah: remaining active paths: 3
Jan 2 17:47:38 server multipathd: mpathas: sdcm - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathas: remaining active paths: 2
Jan 2 17:47:38 server multipathd: mpathw: sdfd - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathw: remaining active paths: 3
Jan 2 17:47:38 server multipathd: mpathah: sdfg - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathah: remaining active paths: 4
Jan 2 17:47:38 server multipathd: mpathas: sdfj - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathas: remaining active paths: 3
Jan 2 17:47:38 server multipathd: mpathw: sdia - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathw: remaining active paths: 4
Jan 2 17:47:38 server multipathd: mpathas: sdig - tur checker reports
path is up
Jan 2 17:47:38 server multipathd: mpathas: remaining active paths: 4
Jan 3 14:25:23 server multipathd: mpathw: remaining active paths: 3

and:

===== paths list =====
uuid hcil dev dev_t pri dm_st chk_st vend/prod/rev

11:0:0:1 sdd 8:48 -1 undef undef XtremIO ,XtremApp
11:0:0:10 sdae 65:224 -1 undef undef XtremIO ,XtremApp
11:0:0:11 sdaj 66:48 -1 undef undef XtremIO ,XtremApp
11:0:0:12 sdal 66:80 -1 undef undef XtremIO ,XtremApp
11:0:0:13 sdan 66:112 -1 undef undef XtremIO ,XtremApp
11:0:0:14 sdap 66:144 -1 undef undef XtremIO ,XtremApp
11:0:0:15 sdaq 66:160 -1 undef undef XtremIO ,XtremApp
11:0:0:16 sdas 66:192 -1 undef undef XtremIO ,XtremApp
11:0:0:17 sdav 66:240 -1 undef undef XtremIO ,XtremApp
11:0:0:18 sdaw 67:0 -1 undef undef XtremIO ,XtremApp
11:0:0:19 sdaz 67:48 -1 undef undef XtremIO ,XtremApp

The system, however works with no multipath or storage errors when booted in a UEK 3 kernel, such as kernel version: 3.8.13-98.7.1.el7uek.x86_64

The following errors may also be logged in /var/log/messages for each failing path:

Jan 2 20:55:52 server multipathd: sdd: unusable path
Jan 2 20:55:52 server multipathd: sdca: unusable path
Jan 2 26 20:56 server kernel: blk_cloned_rq_check_limits: over max size limit.
Jan 2 20:55:56 server kernel: rq_sectors = 512 max_sectors =128
Jan 2 20:55:56 server kernel: blk_cloned_rq_check_limits: over max size limit

and in RAC start up logs, the following error is seen:

ocssd.bin[60996] AFD:IO Error: reqAddr=00000000016f5ce8 reqStart=458768
reqSize=512 on maj/min=249/28 status=0x2e err=-5

Changes

 The kernel was upgraded from the UEK 3 kernel to the UEK 4 kernel

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.