Sun Fire[TM] V1280, E2900, Netra[TM] 1280, and 1290 Servers: IB_SSC fails POST after "Not enough memory" message

Asset ID:	1-72-1269395.1
Update Date:	2016-07-01
Keywords:

Solution Type Problem Resolution Sure

Solution 1269395.1 : Sun Fire[TM] V1280, E2900, Netra[TM] 1280, and 1290 Servers: IB_SSC fails POST after "Not enough memory" message

Related Items


Sun Fire E2900 Server
 Sun Fire V1280 Server
 Sun Netra 1280 Server
 Sun Netra 1290 Server

Related Categories


PLA-Support>Sun Systems>SPARC>Enterprise>SN-SPARC: SF-Exxk
 _Old GCS Categories>Sun Microsystems>Servers>Midrange Servers
 _Old GCS Categories>Sun Microsystems>Servers>Midrange V and Netra Servers

When POST encounters "Not enough memory to allocate buffer of XXXXXX bytes." the next "Excluded unusable, unlicensed, failed or disabled board: /N0/IB6" may be just a side effect that an SC reboot will fix.

In this Document

Symptoms

Changes

Cause

Solution

Applies to:

Sun Fire V1280 Server - Version Not Applicable and later
Sun Fire E2900 Server - Version Not Applicable to Not Applicable [Release N/A]
Sun Netra 1280 Server - Version Not Applicable to Not Applicable [Release N/A]
Sun Netra 1290 Server - Version Not Applicable to Not Applicable [Release N/A]
Information in this document applies to any platform.

Symptoms

The "showlogs" command on the SC or the domain's "lw8" messages show these log entries:

11/14/10 2:46:39 AM Domain Reboot A: Initiating keyswitch: on, domain A.
11/14/10 2:46:39 AM Domain Reboot A: Initiating keyswitch: on, domain A.
11/14/10 2:47:16 AM Not enough memory to allocate buffer of 431300 bytes.
11/14/10 2:47:22 AM Excluded unusable, unlicensed, failed or disabled board: /N0/IB6
11/14/10 2:47:22 AM No usable Io board in domain.
11/14/10 2:47:22 AM Reboot failed.
11/14/10 2:51:20 AM Domain Shell - A: poweroff: Initiating keyswitch: off, domain A.

The next restarts will exclude /N0/IB6, too, making the system unusable.

......
{/N0/IB6/P0} Unknown
{/N0/IB6/P1} Unknown
......

Changes

This happens typically after a long uptime of the SC and frequent reboots of the domain.

Cause

The SC's memory is too fragmented to allocate the required amount of memory.
This leads to POST failing while the IB_SSC is tested.
As a consequence the I/O part, /N0/IB6, is disabled.

There is no formula to calculate the time when the memory will be too fragmented since many variables enter in there.

Solution

Reboot the SC to clear the memory fragmentation.

Use:

lom> resetsc

Because the degree of fragmentation cannot be determined, there is no recommendation when to reboot the SC proactively.

Some guidances are present in the
Sun Fire Midframe & Entry-Level Servers Best Practices Update for Firmware 5.20.x document
http://www.oracle.com/technetwork/articles/systems-hardware-architecture/server-firmware-5-20-x-163882.pdf

Attachments

This solution has no attachment