C6100 BIOS setting from userspace

Wayne_Weilnau at Dell.com Wayne_Weilnau at Dell.com
Tue Mar 20 10:53:42 CDT 2012


Note that OMSA is not officially supported on the C6100.  This may or may not work.  This should be attempted in a lab environment before using it on production servers.

Wayne Weilnau
Systems Management Technologist
Dell | OpenManage Software Development

Please consider the environment before printing this email.

Confidentiality Notice | This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential or proprietary information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, immediately contact the sender by reply e-mail and destroy all copies of the original message.


-----Original Message-----
From: linux-poweredge-bounces-Lists On Behalf Of R, Chandrasekhar
Sent: Tuesday, March 20, 2012 10:50 AM
To: linux-poweredge-Lists; m.j.harvey at imperial.ac.uk
Subject: C6100 BIOS setting from userspace

Hi Harvey,

You can configure the BIOS C-states using omconfig command. Assuming that server BIOS have these capabilities.

The command is as follows: omconfig chassis pwrmanagement config=profile profile=<maxperformance|osctrl|apc|custom>
You can see the command help using "omconfig chassis pwrmanagement -? "

Please refer the OMSA Command line user guide for more details: http://support.dell.com/support/edocs/software/svradmin/7.0/en/CLI/PDF/CLIUG.pdf

Thanks,
Chandrasekhar R
Dell | OpenManage
office +91 80 41178649

Message: 3
Date: Mon, 19 Mar 2012 12:28:44 -0600
From: Stephen Dowdy <sdowdy at ucar.edu>
Subject: Re: C6100 BIOS setting from userspace
To: "Harvey, M J" <m.j.harvey at imperial.ac.uk>,
        "linux-poweredge at dell.com" <linux-poweredge at dell.com>
Message-ID: <4F677ADC.8090102 at ucar.edu>
Content-Type: text/plain; charset=ISO-8859-1

Harvey, M J wrote, On 03/19/2012 11:46 AM:
> Hi,
>
> Hopefully the Poweredge list is also the right place to ask about DCS systems- apologies if not.
>
> I've a bunch of C6100s that I would dearly love to enable C-states on,
> but  the omconfig biossetup tool appears not to support these units.
> The best I've got from tech support is that "it can be set via IPMI", but have drawn a blank on getting actual instructions on how to do this.
>
> Does anyone know how  to set the BIOS C-states (and other) settings from Linux userspace?

Hmm, i'd be surprised if you could change that via IPMI, but please let us know if you find a way to do it.  Perhaps DTK's syscfg has the ability to alter this?

Well, doesn't look like there's a C-State token that can be toggled via 'libsmbios': (at least for this particular system that i don't manage, so i can't just upgrade the BIOS and see if a newer BIOS presents a C-States token interface, but you may have different
results...)

[root at node1 opt]# dmidecode -s system-product-name
C6100
[root at node1 opt]# dmidecode -s bios-version
1.52.00
(that looks like it might be an old BIOS rev)

[root at node1 opt]# /opt/libsmbios/sbin/smbios-token-ctl | egrep -i -C3 '(c|p) state'
================================================================================
  Token: 0x0221 - Power/Performance Management (OS/Hypervisor DBPM)
  value: bool = false
   Desc: Allows OS to change the P state.
================================================================================
  Token: 0x0222 - Power/Performance Management (Dell DBPM)
  value: bool = false
   Desc: BIOS based P state control.
================================================================================
  Token: 0x0223 - Power/Performance Management (None)
  value: bool = true


Whereas on a PowerEdge R710 you see:
foo:~# /opt/libsmbios/sbin/smbios-token-ctl | egrep -i -C3 '(c|p) state'
   Desc: This field controls the number of enabled cores in each processor. By de
         fault, the maximum number of cores per processor will be enabled.
================================================================================
  Token: 0x024b - C States Control (Enabled)
  value: bool = true
   Desc: When set to Enabled, the processor can operate in all availabe Power C S
         tates
================================================================================
  Token: 0x024c - C States Control (Disabled)
  value: bool = false
   Desc: When set to Disabled, there are no C states available for the processor ================================================================================
  Token: 0x0304 - unknown (unknown)
  value: bool = false

which you could change via:

smbios-token-ctl --activate -i 0x024c   (disable C-States)
smbios-token-ctl --activate -i 0x024b   (enable C-States)

--stephen
--
Stephen Dowdy  -  Systems Administrator  -  NCAR/RAL
303.497.2869   -  sdowdy at ucar.edu        -  http://www.ral.ucar.edu/~sdowdy/



------------------------------

Message: 4
Date: Tue, 20 Mar 2012 00:13:05 +0100
From: "gandi.net" <beorn at gandi.net>
Subject: Re: C6100 BIOS setting from userspace
To: linux-poweredge at dell.com
Message-ID: <4F67BD81.20000 at gandi.net>
Content-Type: text/plain; charset=UTF-8

On 19/03/2012 18:46, Harvey, M J wrote:
> Hi,
>
> Hopefully the Poweredge list is also the right place to ask about DCS systems- apologies if not.
>
> I've a bunch of C6100s that I would dearly love to enable C-states on,
> but  the omconfig biossetup tool appears not to support these units.
> The best I've got from tech support is that "it can be set via IPMI", but have drawn a blank on getting actual instructions on how to do this.
>
> Does anyone know how  to set the BIOS C-states (and other) settings from Linux userspace?
>
> Cheers,
>
> Matt
>
>
Hi Matt,

You might want to give a try to http://www.poweredgec.com for your c6100s. It is a good basis for a fully automated poweredge C hardware/firmware management service.


Aur?lien



------------------------------

Message: 5
Date: Tue, 20 Mar 2012 11:18:04 +0000
From: "Harvey, M J" <m.j.harvey at imperial.ac.uk>
Subject: RE: C6100 BIOS setting from userspace
To: Govind Songara <govind.songara at rhul.ac.uk>, Stephen Dowdy
        <sdowdy at ucar.edu>
Cc: "linux-poweredge at dell.com" <linux-poweredge at dell.com>
Message-ID:
        <5E3D0CB98B4EAF41AE2BA04595247FDD473919CF at icexch-m1.ic.ac.uk>
Content-Type: text/plain; charset="us-ascii"

Stephen,

Thanks for that - smbios-tok-ctl  doesn't play completely nicely with the C6100, but it's enough to get P-states enabled.

Govind:  seems that smbios-token-ctl can set cores/CPU and HT on my C6100s (with BIOS 1.67)

Cheers,

Matt

________________________________________
De: Govind Songara [govind.songara at rhul.ac.uk] Enviado el: lunes, 19 de marzo de 2012 22:15
Para: Stephen Dowdy
CC: Harvey, M J; linux-poweredge at dell.com
Asunto: Re: C6100 BIOS setting from userspace

Hi,

I also having couple 10 box of C6100 and want to enable hyper-threading.
Can I set using omconfig?

I am also looking for bios tool by which I can flash the bios to image. And boot the nodes with this image. Does anyone know such tool for Dell bios?

Thanks
Govind


On Mon, Mar 19, 2012 at 6:28 PM, Stephen Dowdy <sdowdy at ucar.edu<mailto:sdowdy at ucar.edu>> wrote:
Harvey, M J wrote, On 03/19/2012 11:46 AM:
> Hi,
>
> Hopefully the Poweredge list is also the right place to ask about DCS systems- apologies if not.
>
> I've a bunch of C6100s that I would dearly love to enable C-states on,
> but  the omconfig biossetup tool appears not to support these units.
> The best I've got from tech support is that "it can be set via IPMI", but have drawn a blank on getting actual instructions on how to do this.
>
> Does anyone know how  to set the BIOS C-states (and other) settings from Linux userspace?

Hmm, i'd be surprised if you could change that via IPMI, but please let us know if you find a way to do it.  Perhaps DTK's syscfg has the ability to alter this?

Well, doesn't look like there's a C-State token that can be toggled via 'libsmbios': (at least for this particular system that i don't manage, so i can't just upgrade the BIOS and see if a newer BIOS presents a C-States token interface, but you may have different
results...)

[root at node1 opt]# dmidecode -s system-product-name
C6100
[root at node1 opt]# dmidecode -s bios-version
1.52.00
(that looks like it might be an old BIOS rev)

[root at node1 opt]# /opt/libsmbios/sbin/smbios-token-ctl | egrep -i -C3 '(c|p) state'
================================================================================
 Token: 0x0221 - Power/Performance Management (OS/Hypervisor DBPM)
 value: bool = false
  Desc: Allows OS to change the P state.
================================================================================
 Token: 0x0222 - Power/Performance Management (Dell DBPM)
 value: bool = false
  Desc: BIOS based P state control.
================================================================================
 Token: 0x0223 - Power/Performance Management (None)
 value: bool = true


Whereas on a PowerEdge R710 you see:
foo:~# /opt/libsmbios/sbin/smbios-token-ctl | egrep -i -C3 '(c|p) state'
  Desc: This field controls the number of enabled cores in each processor. By de
        fault, the maximum number of cores per processor will be enabled.
================================================================================
 Token: 0x024b - C States Control (Enabled)
 value: bool = true
  Desc: When set to Enabled, the processor can operate in all availabe Power C S
        tates
================================================================================
 Token: 0x024c - C States Control (Disabled)
 value: bool = false
  Desc: When set to Disabled, there are no C states available for the processor ================================================================================
 Token: 0x0304 - unknown (unknown)
 value: bool = false

which you could change via:

smbios-token-ctl --activate -i 0x024c   (disable C-States)
smbios-token-ctl --activate -i 0x024b   (enable C-States)

--stephen
--
Stephen Dowdy  -  Systems Administrator  -  NCAR/RAL
303.497.2869   -  sdowdy at ucar.edu<mailto:sdowdy at ucar.edu>        -  http://www.ral.ucar.edu/~sdowdy/<http://www.ral.ucar.edu/%7Esdowdy/>

_______________________________________________
Linux-PowerEdge mailing list
Linux-PowerEdge at dell.com<mailto:Linux-PowerEdge at dell.com>
https://lists.us.dell.com/mailman/listinfo/linux-poweredge



------------------------------

Message: 6
Date: Tue, 20 Mar 2012 16:22:46 +0100 (CET)
From: G.Bakalarski at icm.edu.pl
Subject: linux + Xen + bnx2 +  bonding
To: linux-poweredge at dell.com
Message-ID:
        <773fdbd89cd54ac642d4316956fad7f4.squirrel at webmail.icm.edu.pl>
Content-Type: text/plain;charset=iso-8859-2

Dear ALL.

We have strange problems using bonding module on our DELL R815s farm.

Hardware:

Stacked 2 Juniper Ex-4200 switches (from DELL ;) ) Bunch of R815s - 2 1GBit ports connected (1 to each physical switch enclosure) The R815s have Broadcom 5709 NICs.


Software:

Xen 4.1.2
Linux with kernel 3.2.0 (2.6.32-5-adm64-xen from Debian also tested) In dom0 we have bridges on VLANs and on top bonded interfaces with virtal interfaces for domUs.

Example:
#> ip a

 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
16: vif-pub-dom1: <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master eth-pub state UP qlen 32
    link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fcff:ffff:feff:ffff/64 scope link
       valid_lft forever preferred_lft forever
17: vif-mon-dom1: <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master eth-mon state UP qlen 32
    link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff
    inet6 fe80::fcff:ffff:feff:ffff/64 scope link
       valid_lft forever preferred_lft forever
42: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::16fe:b5ff:feca:4ed5/64 scope link
       valid_lft forever preferred_lft forever
44: eth-pub: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
    inet6 2001:6a0:0:21::d0:20/64 scope global
       valid_lft forever preferred_lft forever
    inet6 fe80::16fe:b5ff:feca:4ed5/64 scope link
       valid_lft forever preferred_lft forever
45: bond0.21 at bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master eth-pub state UP
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::16fe:b5ff:feca:4ed5/64 scope link
       valid_lft forever preferred_lft forever
46: eth-mon: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
    inet6 2001:6a0:1021::2:2000/112 scope global
       valid_lft forever preferred_lft forever
    inet6 fe80::16fe:b5ff:feca:4ed5/64 scope link
       valid_lft forever preferred_lft forever
47: bond0.402 at bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue master eth-mon state UP
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::16fe:b5ff:feca:4ed5/64 scope link
       valid_lft forever preferred_lft forever
48: eth0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP qlen 1000
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
49: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP qlen 1000
    link/ether 14:fe:b5:ca:4e:d5 brd ff:ff:ff:ff:ff:ff
50: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
    link/ether 14:fe:b5:ca:4e:d9 brd ff:ff:ff:ff:ff:ff
51: eth3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN qlen 1000
    link/ether 14:fe:b5:ca:4e:db brd ff:ff:ff:ff:ff:ff


--------------------------------------------------------

bridges:
#> brctl show
bridge name     bridge id               STP enabled     interfaces
eth-mon         8000.14feb5ca4ed5       no              bond0.402
                                                        vif-mon-dom1
eth-pub         8000.14feb5ca4ed5       no              bond0.21
                                                        vif-pub-dom1

----------------------------------------------------------

bonding details:


#> cat /proc/net/bonding/bond0
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation Transmit Hash Policy: layer3+4 (1) MII Status: up MII Polling Interval (ms): 100 Up Delay (ms): 0 Down Delay (ms): 0

802.3ad info
LACP rate: fast
Min links: 0
Aggregator selection policy (ad_select): stable Active Aggregator Info:
        Aggregator ID: 3
        Number of ports: 2
        Actor Key: 17
        Partner Key: 21
        Partner Mac Address: 2c:21:72:9e:b0:80

Slave Interface: eth0
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 14:fe:b5:ca:4e:d5
Aggregator ID: 3
Slave queue ID: 0

Slave Interface: eth1
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 14:fe:b5:ca:4e:d7
Aggregator ID: 3
Slave queue ID: 0
--------------------------------

switch is configured with link aggregaation + LACP

---------------------------------------------------
in dmesg we can see:


[617897.820090] ------------[ cut here ]------------ [617897.820106] WARNING: at
/mnt/linux-2.6-3.2.6/debian/build/source_amd64_none/net/sched/sch_generic.c:255
dev_watchdog+0xe9/0x148()
[617897.820111] Hardware name: PowerEdge R815 [617897.820115] NETDEV WATCHDOG: eth1 (bnx2): transmit queue 0 timed out [617897.820119] Modules linked in: bonding bnx2 xt_physdev xen_netback xen_blkback ebt_ip ebt_ip6 ebtable_filter ebtables bridge xen_evtchn xenfs dm_round_robin dm_multipath scsi_dh ipmi_si ipmi_devintf ipmi_msghandler 8021q garp stp snd_pcm snd_timer snd soundcore snd_page_alloc pcspkr psmouse serio_raw evdev joydev sp5100_tco tpm_tis tpm dcdbas tpm_bios amd64_edac_mod edac_core edac_mce_amd k10temp acpi_power_meter button processor thermal_sys xfs dm_mod nfs lockd fscache auth_rpcgss nfs_acl sunrpc xt_tcpudp xt_state ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4 iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_conntrack nf_defrag_ipv6 ip6table_filter ip6_tables x_tables usbhid hid sg sr_mod cdrom sd_mod ses crc_t10dif enclosure ahci libahci libata lpfc scsi_transport_fc scsi_tgt ohci_hcd ehci_hcd megaraid_sas usbcore usb_common scsi_mod [last unloade
 d: bnx2]
[617897.820257] Pid: 0, comm: swapper/0 Tainted: G        W    3.2.0-1-amd64 #1
[617897.820261] Call Trace:
[617897.820264]  <IRQ>  [<ffffffff81046465>] ? warn_slowpath_common+0x78/0x8c [617897.820283]  [<ffffffff81046511>] ? warn_slowpath_fmt+0x45/0x4a [617897.820290]  [<ffffffff81291f15>] ? netif_tx_lock+0x40/0x72 [617897.820297]  [<ffffffff81292076>] ? dev_watchdog+0xe9/0x148 [617897.820305]  [<ffffffff81051af4>] ? run_timer_softirq+0x19a/0x261 [617897.820311]  [<ffffffff81291f8d>] ? netif_tx_unlock+0x46/0x46 [617897.820318]  [<ffffffff8104ba54>] ? __do_softirq+0xb9/0x177 [617897.820326]  [<ffffffff8120a0ab>] ? __xen_evtchn_do_upcall+0x1b5/0x1f2
[617897.820334]  [<ffffffff8133e8ec>] ? call_softirq+0x1c/0x30 [617897.820342]  [<ffffffff8100f875>] ? do_softirq+0x3c/0x7b [617897.820348]  [<ffffffff8104bcbc>] ? irq_exit+0x3c/0x9a [617897.820354]  [<ffffffff8120b675>] ? xen_evtchn_do_upcall+0x27/0x32 [617897.820360]  [<ffffffff8133e93e>] ? xen_do_hypervisor_callback+0x1e/0x30
[617897.820363]  <EOI>  [<ffffffff810013aa>] ? hypercall_page+0x3aa/0x1000 [617897.820374]  [<ffffffff810013aa>] ? hypercall_page+0x3aa/0x1000 [617897.820382]  [<ffffffff8100663a>] ? xen_safe_halt+0xc/0x13 [617897.820389]  [<ffffffff81014448>] ? default_idle+0x47/0x7f [617897.820395]  [<ffffffff8100d25f>] ? cpu_idle+0xaf/0xf2 [617897.820402]  [<ffffffff81687b38>] ? start_kernel+0x3b8/0x3c3 [617897.820408]  [<ffffffff8168963b>] ? xen_start_kernel+0x586/0x58c [617897.820412] ---[ end trace a7919e7f17c0a757 ]---


and


[617897.820427] bnx2 0000:01:00.1: eth1: DEBUG: intr_sem[0] PCI_CMD[00180006] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: PCI_PM[19002008] PCI_MISC_CFG[92000088] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: EMAC_TX_STATUS[00000008] EMAC_RX_STATUS[00000000] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: RPM_MGMT_PKT_CTRL[40000088] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG:
HC_STATS_INTERRUPT_STATUS[01fe0001]
[617897.824071] bnx2 0000:01:00.1: eth1: <--- start MCP states dump ---> [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: MCP_STATE_P0[0003610e] MCP_STATE_P1[0003610e] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: MCP mode[0000b880] state[80000000] evt_mask[00000500] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: pc[0800c6c8] pc[0800d7d4] instr[ac620038] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: shmem states:
[617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: drv_mb[0103000f] fw_mb[0000000f] link_status[0000006f] drv_pulse_mb[00001d40] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: dev_info_signature[44564903] reset_type[01005254] condition[0003610e] [617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: 000003cc: 44444444 44444444
44444444 00000a28
[617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: 000003dc: 000cffff 00000000
ffff0000 00000000
[617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: 000003ec: 00000000 00000000
00000000 00000000
[617897.824071] bnx2 0000:01:00.1: eth1: DEBUG: 0x3fc[0000ffff] [617897.824071] bnx2 0000:01:00.1: eth1: <--- end MCP states dump ---> [617898.236955] bnx2 0000:01:00.1: eth1: NIC Copper Link is Down [617898.268155] bonding: bond0: link status definitely down for interface eth1, disabling it


The network is very unstable: from timeouts or IPv6 broadcasting somehow filtered/dropped (some hosts not responding to neighbour requests), through one or other VLAN not responding, up to both physical interfaces totally down ....

>From immediate "no connection", through problems after 1.5 hour,  up to
>strange
behaviuor after 5 days ....

Without bonding network was much more stable (i.e.  with only eth0 UP)
- 1-2 months without problmes, however messages in dmesg where also present - without bonding ...

bnx2 version:
2.1.11

The server are not heavily loaded or dont have high network throughput ....

ANY HELP HIGLY APPRECIATED !

GB



------------------------------

_______________________________________________
Linux-PowerEdge mailing list
Linux-PowerEdge at dell.com
https://lists.us.dell.com/mailman/listinfo/linux-poweredge

End of Linux-PowerEdge Digest, Vol 94, Issue 24
***********************************************

_______________________________________________
Linux-PowerEdge mailing list
Linux-PowerEdge at dell.com
https://lists.us.dell.com/mailman/listinfo/linux-poweredge



More information about the Linux-PowerEdge mailing list