Fujitsu SPARC M4000 Product Notes page 33

Specific Issues and Workarounds Concerning Solaris (Continued)
TABLE 7
CR ID Description
6515648 "Replumb Failed" error appears when
dr@0:SB1::memory fails.
6516135 Ap_Id format and devices may not be shown
correctly by cfgadm(1M).
6519290 Large amounts of I/O on swap devices can
cause the system to appear hung by over
welling the I/O system. The amount of I/O
required can be generated through a number
of ways, eg memory shortage, heavy use of
/tmp etc.
6520990 Domain may cause a panic when
deleteboard(8) command for kernel board
by using Dynamic Reconfiguration (DR).
6522017 DR and ZFS may not be used in the same
domain.
Workaround
Once the DR operation is complete, it can be
plumbed up manually.
Example steps to re-plumb the interface
manually:
# ifconfig interface plumb xxx.xxx.xxx.xxx
netmask + broadcast + up
# ifconfig interface group group-name
# ifconfig interface addif xxx.xxx.xxx.xxx
-failover deprecated up
This workaround assumes that the
/etc/hostname.interface file is correctly
configured for the IPMP group and does not
need any modification. The IP addresses used
in the example above should match what was
previously used and what matches the
/etc/hostname.<interface> file.
Use the following operations to display all of
the PCI slots.
1) devfsadm (at Solaris prompt)
2) cfgadm
Set the following to /etc/system and then
reboot the domain:
set maxfastscan=0x2000
To mask this error, add the following entry to
the /etc/system file.
set drmach:fmem_timeout = 30
This bug has been fixed in Solaris 10 8/07.
Reduce the amount of kernel memory that ZFS
can allocate by setting the zfs_arc_max
parameter in the /etc/system file. The
following example sets the maximum size to
512 Mbytes.
set zfs_arc_max = 0x20000000
SPARC Enterprise M4000/M5000 Servers Product Notes 23