New IBM 5015 BBU charge issue

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
H

Herngaard

New Member
May 2, 2013
17
0
1
New IBM M5015 BBU charge issue

Hi Guys,

I'm new here, and new to RAID controllers in general, Googled myself to this nice place, which seemed to be the best place on the internet to ask.

My problem is, that my brand new IBM ServeRAID M5015 controller is having some wierd charge issues with the BBU (IBBU08)
It reaches around 70-76% charge and then states that " Battery Charge Complete" i have tried to run it through a few "Relearn" tasks, and it discharges fine and then reaches somewhere in the 70% range again where it completes it's business.

I did get som warnings about battery temperature reaching 50c, the first few days, but that didn't stop it from charging all the way up to 75%. Now with the cooling fixed and the battery temp running in the mid 40c range, it still stops the charge 1/4 the way short of 100%

I have tried taking the BBU off the card, start up the server, shut it down and put i back on to see if that would help the controller "forget" about the charge level, but it didn't help.

Any ideas to what can cause this, and what i can do to fix it?

my setup is as follows:

intel i7 2600K CPU
Intel Desktop Board DZ68BC
32768 MB DDR3 1600Mhz RAM

IBM ServeRAID M5015 controller with 512mb cache and IBBU08 BBU
Intel Pro/1000 PT Quad Port Server adapter

8x WD RE4 7200Rpm 2TB SATA drives in RAID5 (WD2003FYYS)
3x OCZ SSD's running off the motherboards onboard controllers

Fractal Design Define R4 midi-tower
Corsair H100i CPU watercooler (exausts from case through 2x 120mm fans in top)
3x Corsair AP140 Quiet 140mm case fans (two intakes in front, one exaust)
Corsair AX750W Gold PSU

BIOS Version on the controller:
3.27.00_44.12.05.00_0x05270000
Firmware Package version:
12.13.0-0154

OS: Windows Server 2012
MegaRAID Storage Manager - 13.01.04.00 (client build 2.90)

Any help would be greatly appreciated.

- Herngaard
 
Last edited:
M

mobilenvidia

Moderator
Sep 25, 2011
1,956
212
63
New Zealand
Try going back to an older FW, possibly an IBM one might be best
Then do a relearn
See if this helps.
If it does then go forward in FW again.

But you might have a BBU with a faulty cell, it will charge 3/4 of the cells but the other(s) might not take any charge.
 
H

Herngaard

New Member
May 2, 2013
17
0
1
thanks mobilenvidia,

When i first installed the card, i had the server up and down several times messing around with harddrives and cables, so i suspect that the initial battery learning process was messed up somehow. I flashed to the LSI 9260-8i BIOS (that i heard was supported) because i cant get to IBM's own:

 
H

Herngaard

New Member
May 2, 2013
17
0
1
Can anyone help me with a link to the most recent firmware made by IBM?
I just tried a relearn process again, and it stopped at 72% after completely discharging. Restarting the server would add another 2 percent, as it seems to charge a bit after every startup, a couple restarts more added 1 percent, then stops there.

I'm wondering if i should just return it for another one, but thinkin i should at least give the original firmware a change first.
 
R

ruffy91

Member
Oct 6, 2012
71
11
8
Switzerland
My M5014s BBU also stays at 74% charge, I don't think this is a problem. IMHO it charges only to 48Hrs+ retention time to conserve battery lifetime (80% is ideal charge for Li-Ion/Li-Po batteries, 100% shortens lifetime)
 
H

Herngaard

New Member
May 2, 2013
17
0
1
ruffy91 you might be right, it´s supposed to reach a 48 hour renention time, and it does exactly that at 74%


It would be interresting to see if everyone else with 5014 and 5015's and IBBU08 batteries are experienciing the same.

It does look a bit "messy" to me :)
 
H

Herngaard

New Member
May 2, 2013
17
0
1
I found this about charge cycle behavior:

ServeRAID-M Series Battery Backup Unit (BBU) charge cycle behavior and cache modes - IBM System x

It is not an uncommon event for the "Relative state of charge" of the BBU not to reach 100%. Relative state of charge is an indication of full charge capacity percentage in relation to the design capacity. Multiple battery learn cycles may move this state upward or downward.
Click to expand...
Also found these values for the different modes:


I changed mine to mode 1: which gives me an estimated battery life of 5 years.

Full document here:
Support

Glad i didn't send it back :)

EDIT:
BTW, Right now i'm running a relearn task to se what percentage it ends up at, with the low 12H retention time, which only requires 400 mah compared to the 960 mah of my previous setting.
I'll post back with my findings.
 
Last edited:
H

Herngaard

New Member
May 2, 2013
17
0
1
Ok, so after changing the retention mode to mode 1 (in the webbios via ctrl+h) and restarting the server. I chose to do a relearn.

It didnt respond to the task, so i restarted again and it started charging the battery all the way to 100% :)


Right now it discharging, so cant wait to see what it ends up at once it's done with that.

 
M

mrkrad

Well-Known Member
Oct 13, 2012
1,244
52
48
i take the batteries out of LSI controllers (SSD), they cause more problems than not. seriously. Almost every time the M5014's have trouble with batteries, issues start cropping up. same with the H700, if you want battery back - use supercapacitor or use a controller like HP which doesn't have cycles
 
H

Herngaard

New Member
May 2, 2013
17
0
1
With a retention time of 12 hours the charge stopped at 27% (the reason it's 31 is because it charges for a couple of minutes at reboot)

 
M

mrkrad

Well-Known Member
Oct 13, 2012
1,244
52
48
just wait for it to start causing bus resets :) you know if this is in a server like 1U you need to move the battery to another area, otherwise the heatsink will not flow and the card will overheat and reset.

if you don't have a pair of fans blowing on the card, best not have that battery directly mounted. 45C is pretty hot for batteries, heat kills batteries. no joke man.
 
H

Herngaard

New Member
May 2, 2013
17
0
1
I was thinking the same, it's getting pretty hot. I have two 140mm high airflow fans in the front of the case that i would much rather have the battery in front of.

I cant really find much info about that cable, do you know if the 20pin cable like this one is all i need:
 
M

mrkrad

Well-Known Member
Oct 13, 2012
1,244
52
48
well they sell a $20 card to mount it, but you can do something similar, I'd suggest using their card if you have an open bay so you get the LFM airflow that a server can provide.

I had these old tripp lite top heavy rectangle ups, they ran at 130-140F. I replaced the battery on two, both spit flames out the front - so of course I bought two replacement units since the batteries cost a lot. From firebuy.com - lol. One of those spit fire after shutting it down and trying to move it.

I chucked them and went back to APC only.

With supercap technology there is no reason on the planet to stick to batteries. They are obsolete.
 
Last edited:
H

Herngaard

New Member
May 2, 2013
17
0
1
I like the idea of a cheap UPS instead. I'll look into one of those APC ones, they seem to be very common.
What have changed with supercap technology? UPS's have been around forever?
 
M

mrkrad

Well-Known Member
Oct 13, 2012
1,244
52
48
the supercap is totally different design. When power is lost , there is a fraction of time the capacitor can hold the flash backup nand to copy from the controller ram cache.

battery backed write cache just keeps the ram "refreshed". One needs only enough power to transfer the active ram to the flash board, the other keeps the memory going until you can power up the system.

It's likely a supercap could survive a small crisis where power may be out for a week, it's not likely battery back write cache could run long. Typically the batteries are 72hours maximum when new. Dell/LSI are more into the 24-36 hours.

If you don't watch your systems, and on the way home from work friday, until monday AM you get in, that can be a large # of hours eh?
 
H

Herngaard

New Member
May 2, 2013
17
0
1
ok, thanks for the explanation :)
So do the RAID controllers have this NAND cache today, and does that make writeback safe without a battery today?
 
M

mrkrad

Well-Known Member
Oct 13, 2012
1,244
52
48
Supercapacitors look like triple-A batteries. Two for 5.4volts. LSI calls the technology cachevault. HP just calls it Flash Back Write cache.

Anyways, the point is that it takes about a second to charge a capacitor and no "cycles".

IMO, i'd focus on controlled shutdown of your environment (N-WAY) power redundancy.

Bad shit happens if you cut the power on an active server.
 
You must log in or register to reply here.
Top Bottom