Asset ID: |
1-71-1004591.1 |
Update Date: | 2011-01-28 |
Keywords: | |
Solution Type
Technical Instruction Sure
Solution
1004591.1
:
PICLD Overview and Fixes To Common Bugs
Related Items |
- Sun Blade 2000 Workstation
- Sun Fire V480 Server
- Sun Fire 280R Server
- Sun Netra 20 Server
- Sun Fire V880 Server
- Solaris SPARC Operating System
- Sun Blade 1000 Workstation
|
Related Categories |
- GCS>Sun Microsystems>Desktops>Workstations
|
PreviouslyPublishedAs
206370
DescriptionThis article is a PICLD Overview and addresses Fixes To Common Bugs.
Steps to FollowWhat is the name of the Environment Monitoring and Control daemon.
picld - PICL daemon
--------------------------------------------------------------------------------
Brief description of PICLD
The Platform Information and Control Library (PICL) provides a mechanism to
publish platform-specific information for clients to access in a
platform-indpendent way. PICLD maintains and controls the PICL information
from clients and plug-in modules. The daemon is started in both single-user and
multi-user bootmode.
--------------------------------------------------------------------------------
The first version picld is included in: Solaris[TM] 8
--------------------------------------------------------------------------------
Platforms supported on are:
Sun Fire[TM] 280R
Sun Fire[TM] 480R
Sun Fire[TM] 490
Sun Fire[TM] 880
Sun Fire[TM] 890
Sun Blade[TM] 1000
Netra[TM] T4 aka Netra[TM] 20
Netra[TM] T12 aka Netra[TM] 1280
Sun Fire[TM] V1280
Sun Fire[TM] E2900
Sun Fire[TM] 4800
Sun Fire[TM] 4900
Sun Fire[TM] 6800
Sun Fire[TM] 6900
Sun Fire[TM] 12K/15K
Sun Fire[TM] 20K/25K
--------------------------------------------------------------------------------
Files:
PICL daemon door: /var/run/picld_door
PICL daemon: /usr/lib/picl/picld
Start/stop script: /etc/init.d/picld
--------------------------------------------------------------------------------
Is the PICL daemon necessary for system use or can it be shutdown without
impacting functionality
Although the system would not have the ability to monitor system events
the system would continue to run. Monitoring of all removal and insertion
activities, temperature ranges exceeding thresholds, fan failures, and power
failures would not be reported at Solaris level.
If the picld daemon fails to start, then environment daemon (piclenvd)
won't be running and CPU temperatures won't be monitored. This can result
in CPU overheating. In addition to this, there are a number of other subsystems
that may depend on this daemon. /usr/sbin/prtdiag is just one example of this.
--------------------------------------------------------------------------------
Does 'picld 'daemon send messages to system console and
/var/adm/messages file
Yes, the 'picld' deamon provides output to the system console and the
/var/adm/messages file when certain events are detected. These events
include
removal and insertion, temperature & fault conditions.
Note - When replacing a fan tray or power supply, insert the new device after
a console message displays indicating the old device has been removed.
Caution - The software polling process takes up to 30 seconds to recognize
device insertion or removal, so allow sufficient time before inserting
a device into a recently cleared slot.
Note - Replace the primary fan trays before the secondary, especially in
the case of the CPU fan trays. The primary CPU fan tray operates at a
variable speed and is used to control the temperature of the CPUs.
===============================================================================
Here are some of the known bugs commonly encountered:
Following errors are reported in /var/adm/messages after
OS installation is complete on SUNW,Sun-Fire-280R, Bug-ID# 4458436
May 17 09:05:30 d34c picld[69]: [ID 299567 daemon.error] No FRU Information
for
CPU0_DIE_TEMPERATURE_SENSOR using default temperatures May 17
09:05:30 d34c
picld[69]: [ID 478985 daemon.error] ERROR running
psvc_update_thresholds_0 on
CPU1_DIE_TEMPERATURE_SENSOR (1604840) May 17 09:05:30 d34c
picld[69]: [ID 875627
daemon.error] No such file or directory
Suggested Fix: apply Patch-ID# 110460-08
--------------------------------------------------------------------------------
picld will sometimes display errors after power supply hot plug.
Messages from console output, Bug-ID# 4431165
Mar 28 14:07:46 wgs48-58 picld[70]: Device PS0 removed Mar 28 14:07:51
wgs48-58
picld[70]: Device PS0 inserted Mar 28 14:07:58 wgs48-58 picld[70]: Device
PS0
removed Mar 28 14:08:07 wgs48-58 picld[70]: ERROR running
psvc_ps_device_fail_notifier_policy_0 on PS0 (2361480) Mar 28 14:08:07
wgs48-58
picld[70]: No such device or address
Suggested Fix: apply Patch-ID# 110849-04
--------------------------------------------------------------------------------
picld daemon errors in /var/adm/messages with >1GB RAM is installed:
Bug-ID#s 4432412 and 4451949
Mar 29 15:45:40 option3 picld[88]: [ID 325715 daemon.error] SUNW_picl
memcfg
physical memory tree failed! Mar 29 15:47:05 option3 pseudo: [ID 129642
kern.info] pseudo-device: devinfo0 Mar 29 15:47:05 option3 genunix: [ID
936769
kern.info] devinfo0 is /pseudo/devinfo@0
Suggested Fix: apply Patch-ID# 110460-01
--------------------------------------------------------------------------------
The following errors are sometimes seen when inserting a power supply:
Bug-ID#s 4413285, 4414411, and 4356073
Feb 7 14:11:02 wgs48-76 picld[77]: Device PS0 removed Feb 7 14:11:12
wgs48-76
picld[77]: ERROR running psvc_ps_device_fail_notifier_policy_0 on PS0 Feb
7
14:11:12 wgs48-76 picld[77]: No such device or address Feb 7 14:11:12
wgs48-76
picld[77]: ERROR running psvc_ps_overcurrent_check_policy_0 on SYSTEM
Suggested Fix: apply Patch-ID# 111792-17
--------------------------------------------------------------------------------
In addition picld daemon may encounter memory issues:
Bug-ID#s 4431402 and 4417600
Sending lots of SIGHUP signal to picld daemon causes picld to grow over time,
indicating some kind of memory leak.
Suggested Fix: apply Patch-ID# 108528
--------------------------------------------------------------------------------
Why is picld using 1.3 % of CPU and 57% of memory
Bug-ID# 4515266
USER PID %CPU %MEM SZ RSS TT S START TIME COMMAND
root 195 1.3 57.2 122832722282528 S Jan 07 2027:11
The first is to restart picld before the leak causes problems on the system
or messages on the console.
If the picld process ended running out of memory every week you could simply
restart picld before that time and you would never see the out of memory
error. To do this you simply run...
#/etc/init.d/picld stop
#/etc/init.d/picld start
Suggested Fix: apply Patch-ID# 110849-09
--------------------------------------------------------------------------------
picld uses fork() to restart itself, which may cause deadlock on certain
memory corruption errors:
Bug-ID# 4459534
picld restarts itself in subnormal mode when unexpected errors are
encountered
during its execution. It starts up in degraded mode first, and then in failsafe
mode, before exiting. The main reason for doing this is to allow critical
functionality, like environmental monitoring, to continue.
Suggested Fix: apply Patch-ID# 108528
--------------------------------------------------------------------------------
picld alignment panic occurred during DR test libcfgadm_031_040:
Bug-ID# 4401168
panic[cpu8]/thread=30018e83980: BAD TRAP: type=34 rp=f0472d40
addr=6969696969696979 mmu_fsr=0 picld: alignment error:
addr=0x6969696969696979
pid=100070, pc=0xf003efb4, sp=0xf0472601, tstate=0x0, context=0x1801
o0-o7:
f0801e60, f0826ba0, 0, f0826e60, 300183fa840, 30000102000, f0472601,
f002f620
g1-g7: 0, 0, 101a3b98, f0801e60, 186a0, 10471778,
104717a200000000f0472a50
unix:die+b0 (34, f0472d40, 6969696969696979, 0,f0472d40,10432ed0 %l0-3:
0000042c00001000 00000c6800000e29 0000000000000000
0000000000000001 %l4-7:
0000000000000000 0000000000000001 0000000000000002
00000000f08067f0
00000000f0472b30 unix:trap+64c (0, 0, 10000, 0, f0472d40, 0) %l0-3:
0000000000010200 0000030018c98ac0 0000000000000000 0000000000007fe
%l4-7:
0080000900000034 0000030018e87530 0000000000800009
00000000104bda70
00000000f0472c90 unix:user_rtt+32c (2a1006517fc, 9, 0, 2a100651738, a,0)
%l0-3:
0000000000000006 0000000000001400 0000000080001605
0000000010037040 %l4-7:
0000000000000000 0000000000000000 0000000000000000
00000000f0472d40
00000000f0472de0 f003ef90 (0, 1f, 16, 3c, 0, f08296c0) %l0-3:
0000000000000000
0000000000000000 0000000000000005 00000000f043922 %l4-7:
0000000000000000
0000000000000000 0000000000000000 0000000000000001 panic: entering
debugger (continue to save dump) Type 'go' to resume debugger entered.
Suggested Fix: apply Patch-ID# 110918-03 or latest KJP 108528-15
ProductSolaris
Sun Fire V480 Server
Sun Fire V880 Server
Sun Fire 280R Server
Sun Blade 2000 Workstation
Sun Blade 1000 Workstation
Netra 20 Server
picld
Previously Published As
45748
Change History
Date: 2010-12-21
User name: Dencho Kojucharov
Action: Currency check
Comments: audited by Entry-Level SPARC Content Lead
This document applies only to Solaris 8, so no further edits are needed.
"MIGRATION CLEANUP"
The Product "Solaris" is too broad. In order to publish this document, all
applicable versions of Solaris must be entered in the product statement (i.e.
Solaris 9 operating system, Solaris 8 Operating System...etc). You can look up
the possibilities using the Swordfish Lookup tool (gives proper product
Nomenclature). http://krep.emea.sun.com/stats/swordfish/
Search Solaris and after clicking on applicable versions click Result button and
copy and paste into product statement.
Product_uuid
3285bfa4-224e-11d6-8eb3-843d3a923213|Solaris
a2b9bc2b-52c6-45c2-a3e0-f19bd2c86953|Sun Fire V480 Server
29726712-0a18-11d6-8636-c7e996b581dc|Sun Fire V880 Server
296f2476-0a18-11d6-86cf-c8096baa086c|Sun Fire 280R Server
0be55d30-2c89-4d9b-8027-344c808940d9|Sun Blade 2000 Workstation
2a0b1052-0a18-11d6-9854-d945df5eb308|Sun Blade 1000 Workstation
29aa2760-0a18-11d6-9ba6-ecdc03b9c8f5|Netra 20 Server
Attachments
This solution has no attachment