Host battery status alarm

[SirOracle]

Hi,

I have installed ESXi4 on some servers and they all come up with an alarm in vCenter:

Status: Warning

Name: Host battery status

These are all Dell PowerEdge 2950 servers. Does anyone know what this alarm means? Could this be a bug and should I disable this alarm?

[1905]

Hi

We have the same problem on hardware FUJITSI SIEMENS PRIMERGY BX620 S4.

[TAMW]

We have the same. This is not a bug. Sometime this alarm is triggerd but it always goes back to green after a while (few hours).

From time to time the hardware wants to check te battery of raid BBWC en then you get the waring.

You also get the warning the battery is simply charging.

[SirOracle]

If what you say is correct and this is not a bug, I will disable this alarm. I can't go in every day and check to see if the alarm is just about the battery charging, I will end up ignoring the alarm sign on the hosts and miss a real alarm.

[ufo8mydog]

Hi SirOracle,

We did not have any issues on ESX3.5 (on our PowerEdge 1950's), once we installed Vsphere two of the hosts had this weird phantom alarm that we could not shake.

A polite staffer from VMWare Global support told me that it was something they were aware of internally, so If I were you I would just disable that alarm until further notice.

[SirOracle]

All the servers are back to normal now, but it took several days. Under the tab Hardware Status I saw more information about the warning, "Learn cycle requested", not sure what was meant by that, but it all looks fine now.

Thanks guyes for keeping me calm while I waited for the warning to go away :smileywink:

[jiths]

It is the alert generated from ESX CIM layer for your internal storage Controller BBU (Battery Backup Unit). I never seen this issue much on our servers, so it is not a bad idea to check the Storage Controller battery once.

[jasoncllsystems]

Do you have any Dell OMSA install? Probably you can see more details on OMSA.

Regards,

MALAYSIA VMware Communities

CLL SYSTEMS

#if you found this or any other answer useful please consider allocating points for helpful or correct answers ***

[devzero]

i have this problem with FSC Primergy RX300 S3, but with ESX4 (not ESXi)

Systemstatus shows "Battery Status: Charging" (Leran Cycle Complete) with a yellow sign for more than 24hrs now.

[s1xth]

I have this same issue (have had this since I deployed my first ESXi back on 3.5 U1)...its all ok. Its just the RAID controller battery going through a 'learn' 'recharge' cycle. I wouldnt worry about it. i have all Dell PE2950's and R710s.

[ziolek]

I have this problem with FSC Primergy RX300 S5 with ESX4. It's very infuriating alarm. :smileyhappy:

[estereterete]

I'm having this two warnings:

• Battery on HPSA1,
  
  
• HP Smart Array E200i Controller : Embedded : HPAS1
  
  

in HP BL465 G5 since I updated from ESX 4.1 to hp Esxi 4.1 https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=HPVM06

This warning never gets back to green.  I though it was ESXi, so I hoped it would stop after installing HP ESXi image. Sadly, I still have the same problem.

Did somebody solve this?

[J1mbo]

Most likely the battery is bad.  Check your hardware IPMI logs or similar.

[101Chris]

It does not seem to be a problem with the communication between ESX and the Dell health report agent or whatever they call it as I first thought.

The common denominator between a lot of Dell systems is the PERC6 raid controller. The PERC6 is a re-branded LSI Logic board, which would explain how the same issue appears across different brands of machines.

When reading trough the changelog of the latest firmware upgrades for the PERC6 controller, I found this:

4. _Corrected an issue where the battery learn cycle delay was not delaying for the right amount of time._

So.. that should be it, and was. After I upgraded the PERC6 firmware, my three PE R610 stopped complaining in vCenter.

To everyone with this issue, upgrade firmware on your raid controller :smileyhappy:

//Chris

[s1xth]

Yeah I saw the same thing too in the changelog, that must be the fix. I am assuming you flash the controllers firmware via the Dell Update disc correct?

[101Chris]

Well, acually, with these Gen11 servers Dell has implemented somthing they call USC.(Unified Server Configurator) and it's a part of iDRAC that connects to a repository and downloads whatever you want to update. So basicly i can remote my machines from the office, reboot it and enter the USC and run firmware and driver updates through that... Now if I only could figure out how to get the backup tapes to run them selves over to the bank im set! :-D

But with older machines i suppose u need a bootable media with the fw files on it as wel. (could be wrong tho)

//Chris

[FredericNass]

Hi,

The same upgrade on DELL R610 didn't make it for us.  ?:|

Still thoses 'Host battery status' alerts showing when ESX restores from standby or shutdown.

BTW, thoses alerts are not showing everytime.

Another thing to notice is that  this "4. Corrected an issue where the battery learn cycle delay was not delaying for the right amount of time." fix didn't come with R216024 but with previous version R201071.

Dunno if the latest version did not reintroduced this "bug" and I didn't want to downgrade to verify this. So I'll still observe thoses alarms and if comming too many often, will simply disable them.

This is really anoying. I'll try to get some more infos from DELL & VMWare ASAP.

[Thomy23]

Hi Guys

We have 3 Dell R710 Servers in our Vsphere Cluster:

PERC FW: 6.2.0 (up to date)

We have the problem with all of the 3 servers, even a BIOS Update from 1.1.4 to 1.2.6 didn't solve the problem. The only solution for us was to disable the alarm :smileyblush:

Greets Thomy

[vm-au-user]

I'm getting the same issue across all our IBM 3850M2

[mjpagan]

I have also been getting this issue on every IBM x3650M2 that I've placed in the field. Has anyone fixed it for a IBM or can someone from VMware comment on the issue?

Mike P

MCSE, VCP3/4

[J1mbo]

It is annoying but[/b] as said above the learn cycle is only once every 3 months (and takes about 3 days to complete). The learn cycle is required for the controller to genuinely report when the lithium ion cell is too worn, and at this point you absolutely need to know about that if you are using local storage, since the Perc will drop to write-through mode and parity based RAID write performance will take a massive hit as a result.

But I agree vSphere should differentiate between learning mode and real failure.

Please award points to any useful answer.

[umarzuki]

this is happening on x3650m2 running esx 4.0 update 1 and vcenter 4

[ArildS]

We have this issue with a new Supermicro H8QM3-2 2041M-32R+B server with an LSI Megaraid 8708ELP SAS controller (with BBU).

Seems to be the LSI controller that causes this (about the same as the PERC6 LSI controller rebranded to Dell).

We run the newest LSI firmware (11.0.1.0017) and we still get this error message every other day. Even though the battery learn cycle are complete at the moment. Learn cycle repeat every 90 or 120 days if I remember correctly.

This message should not appear so frequently, so either this is a firmware issue where the firmware sends faulty data to ESX or ESX is reporting an error where no error occurs.

[Bitfarmer]

Well I just verified that upgrading to the latest 6.2.13 firmware for the Perc 6i controller in my Dell 2900's doesn't make the alarm go away. I had two hosts with the error that were running the latest and as soon as I took my 3rd host from 6.1.x to 6.2.13 the alarm immediately showed up on it. After waiting a day and updating the hardware status they went away on their own again. Opened a case with VM support asking why this keeps happening and their answer was basically it's a hardware problem, are you sure there's nothing wrong with your servers. Told them it's a cross platform vendor issue and there's nothing wrong with my servers. So they just basically gave me the blow off and told me to ignore the alarm. There's good practice. I think VM must have issues interpreting the status in general and they need to tweek how the sensor is monitored since it happens across dis-similar hardware.

[FranciscoBatist]

Well guys i'm getting these anoying Host Battery Status Alams on 2 IBM x3650 M3 installed with VMware ESXi 4.1.0 build 381591.

Everything is up to date, and the hardware is fine. In the past i had a similar issue on ESX 3.5 with another IBM machine, that was only solved when we upgraded to ESX4, by the time, i even got confirmation from VMware that this was a know issue that was going to be corrected in future releases.

For me this is the same problem, VMware monitoring for this machine is probably misinterpreting the data gathered.

Any ideas? Anyone with similar issues? Anyone heard from VMware that it's a know issue?

I will open a problem with IBm, but after asking me if everything is upto date (VMware and Firmwares) thay will tell me to speak with VMware.

Best regards.

Will get to you soon.

Francisco Batista - VCP 3 & 4 - VSP - VTSP

Softinsa - Portugal

[FranciscoBatist]

Upgrading to esxi 5 solved the alarm anoying problem for my costumers.

[Hi5]

Hi,

We have 2 Dell PE R710 and 2 PE 2950 with ESX 4.0 U2.

We have been having the same problem for years. We just ignored battery status alarms.

Recently we had a problem with air conditioning in the server room and temperature was about 35 degrees C for about 24 hours.

However when I went to the Hardware status tab on each ESX server the System Board Ambient Temp was 25 degrees. On the top right corner it was showing that status was updated 5 days ago. When I hit Update it changed to 40 degrees.

It looks like it goes to sleep mode sometimes and just do not update data until you do it manually. It makes hardware status alarms unreliable for us.

Did you guys find any solutions?

Thanks

[peetz]

Maybe you are experiencing the issue described here:

vCenter Server Hardware Status shows the battery status of RAID controllers as failed or in question

Andreas

- check out my VMware Front Experience Blog

[estereterete]

Yeah, that's my issue http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1021786

I've had 3 servers working fine for months with ESX4.1 The moment I upgraded them to ESXi 4.1, this alarm started raising..

I've checked, as well, that this alarm is only triggered if esxi has v.m. running. I don't see any relationship, but it's happening from the beginning.

[JohnnyLeung]

Hi Guys

I also have similar problem. I am using IBM X3650 M3 server with esxi 4.1 update1.

We received twice battery warning message within 2 days.

We have verified that there is no error message on the physical server. But keep receiving this warning. Could someone help me to solve this issue, many thanks.

Johnny

[nlandas]

    This is likely an informative message and under the hardware tab there are more details. However, keep in mind that at least on the DELL PE2950, DELL recommends replacing those PERC Raid batteries every three years. You can argue that this is too frequently but if the server is off it's initial 3 year warranty, maybe it is time to replace that PERC battery. It is an important part of the system.

    Now the alarms should not be on a long time if the battery is simply charging/testing/learning/etc. so as others said on the thread this looks like a bug with certain versions and patch levels of ESXi but still woth looking at the age of the battery and recommended replacement cycle.

[J1mbo]

As of build 502767 (and possibly before) the learn cycle is correctly reported at least for Perc-6i controllers, "Battery on Controller 0 Discharging ... (Learn Cycle Active)"

[JohnnyLeung]

Hello guys

Now I have similar problem again "Battery on Controller 0 (Learn Cycle Requested)" for more than 36 hours, I have tried to reboot the server several times, but it does not help.

IBM has confirmed that the battery does not need a replacement, but its in Learn Cycle Requested, anyone knows how long will it take for a learn cycle, and how to determine if the warning message is fake or not and how to get rid of it?

Thanks a lot for your help.

[Nilopher25]

Even I face the same error with my organization's esx servers ... but disabling the alarm is not the only solution .... what if I really want to monitor the battery status and after disabling the alarm , obviously it won't be possible .... Any suggestions or alternatives ???

[SkyGuide]

I had the same issue in my data center. Old Dell PowerEdge 1950 servers started to generate this alarm.

Only the servers with a separate dedicated battery to for RAID controller where generating this alarm.

So in my case this was NOT a ghost alarm, it was very real.

The battery in question is connected to the RAID controller to keep the memory modules alive for RAID while its moving data back and forth between the drives.

This batteries do have a finite life span and eventually they do stop charging.

The PE 1950 without the separate RAID card installed (and no separate battery) did not had this issue. All my PE1950 where made in the same year.

SOLUTION... in my case was to buy the new battery, shut the host down, replace the battery, restart the host, Acknowledge the alarm, reset it, and that's it.

Once I replaced the battery in two of my old PE1950 - alarms stopped.

Hope this helped.

The Battery you want to replace sitting next to the RAID card in the left corner of the server (PE 1950) close to the front panel, near by the power button. Battery kind of large, inclosed in the black plastic, resembling old fat cell phone batteries.