Sun Fire[tm] 3800 Server: Repair Procedures
This section provides the most common repair procedures for the
Sun Fire 3800, which are included in the Sun
Fire 6800/4810/4800/3800 Systems Service Manual (805-7363).
Clustered Hardware Note: When repairing a clustered system,
you should replace server components by first switching over the data services
to the functioning server, halting the host to be serviced, powering down the
host, and then performing the hardware procedure to replace the component. Following
the procedure, the logical hosts should be switched back to the default masters.
The following table provides the page numbers for each procedure.
Component |
Tips |
Centerplane |
- Certain Sun Fire 3800 systems with duplicate MAC addresses running
on common subnets may experience network issues (FIN I0857-1).
|
Compact PCI (cPCI) I/O Assembly |
- Sun Fire Servers (3800/4800/4810/6800) may encounter panic during
Dynamic Reconfiguration (DR) operation of PCI and cPCI I/O boards (FIN
I0840-1).
- Sun Fire x800 systems are subject to a "panic" problem at
boot time when a cPCI Dual FC Network Adapter (FRU or X-option) is installed
for the first time (FIN I0708-1)
- Sunfire 3800 systems with 540-4591-02 (or lower) cPCI I/O Assemblies
will not support Hot Swapping of cPCI cards (FCO A0190-1).
|
Compact PCI (cPCI) Cards |
- Sun Fire x800 systems are subject to a "panic" problem at
boot time when a cPCI Dual FC Network Adapter (FRU or X-option) is installed
for the first time (FIN I0708-1)
|
CPU/Memory Board |
- CPU/Memory Board FRUs for Sun Fire 12K/15K and Sun Fire 3800-6800
systems are not enabled for Capacity On Demand (COD) use (FIN I0912-1).
- Guidelines for understanding and diagnosing UltraSPARC III Level 2
(L2) SRAM Cache Memory Errors (FIN I0887-1).
- UltraSPARC III and III+ based platforms could be susceptible to UCC
errors that may cause system panics (FIN I0856-1).
- Issue with diagnosing of "send mondo" panics (FIN I0765-1)
- 900MHz CPUs operating at 750MHz (FIN I0759-1)
- Under certain conditions, flashupdate -u command can create incompatible
firmware versions between boards, rendering system unusable until problem
is corrected (FIN I0731-1)
- A limited set of Sun Fire System Boards may be vulnerable to Uncorrectable
Errors in L2 SRAM. (FCO A0248-1).
|
Disk Drives |
-
Improved firmware for Seagate 10K.6 disk drives will reduce the incidence of unexpected outages due to a spindle motor issue (FIN I1136-1).
|
Memory |
- On certain Sun Systems a small number of 512MB and 1GB Micron DIMMs may experience premature failures with Uncorrectable Memory Error (UE) messages. (FAB 1000920.1) (formerly FCO A0285-1)
- Best Practices Guide for Memory Errors for diagnosing UltraSPARC III
memory errors now available (FIN I1018-1).
- Diagnosing Main Memory errors versus L2SRAM errors on UltraSPARC III
and UltraSPARC III Cu systems (FIN I0954-1).
- A sub-population of DIMMs that shipped between 2001 and 2002
on the below platforms are showing significantly lower
reliability than expected (FCO A0253-1).
- Systems containing 256MB Samsung B-die DIMMs, having a module date
code between 0115 and 0127 (built between weeks 15 and 27 of 2001),
may experience Uncorrectable Memory Errors (UE). This can lead to System
Panics (FCO A0223-1).
|
System Controller Board |
- Sun Fire 3800/4800/4810/6800, V1280 and Netra 1280 systems may experience
false data parity error messages (FIN I1020-1).
- Sun Fire System Controllers with 5.11.x firmware could experience
a loss of network settings during a firmware upgrade (FIN I0891-1).
- Sun Fire 3800/4800/4810/6800 domains can hang when System Controller
firmware is downgraded from 5.13.0, 5.13.1, or 5.13.2 to 5.12.7 or lower
(FIN I0890-1)
- Systems with redundant SCs may experience failed domains after clock
failure (FIN I0762-1)
- Unexpected behavior from SCs due to downrev firmware (FIN I0756-1)
|
|