HP Proliant ML110 G6 Breaks down offen

[dwesely]

Hey everyone, I found this thread via Google and have been experiencing the exact same problems with my ML110 G6. I contacted HP and as some of you probably have found they try to blame anything but their server. Originally, I had 16 GB of Kingston RAM (passed sevaral memory tests with different utilities) which I ended up having to ditch in favor of HP branded memory so they'll support it which I've installed to no avail. I also had just two hard drives in the beginning using the built in RAID capabilities - still had issues. The server is now running SBS2011 with 4 HP hard drives configured on a P212 (256 MB) in two RAID 1 arrays. All current drivers and firmware yet the thing locks up every two or three weeks. I'm basically the only one using it right now so the load is very low.

One thing I noticed but didn't really think about it too much at the time but I didn't have problems with SBS2011 pre-SP1. There could've been an update that broke something? Odds are probably slim but if anyone has a problem with a non-production ML110 they could try testing for it. Worth a shot.

At this point I've configured iLO to see if I can access that when the server stops responding so next time. Anyone have any luck with the ML110 G6 after disabling memory interleaving?

[DGTJBarker]

Can can confirm that this fix worked for me as well (disabling Memory Interleaving).  This is also on a machine with 10 GB (2GB + 4GB + 4GB).

[Akorostel]

I also have the same problem with two Proliant ML110 G6 servers (each has 2x2 Gb RAM it shipped with). Servers freeze with period from 2 to 20 days. Screen stops updating, no reaction to mouse and keyboard. However, the "frozen" server is available via network: I can ping it, can explore shared folders, etc. I tried to connect to the installed VNC server application. After long wait (about 5 minutes), it eventually asked for password, but failed to display the server screen (I've waited for about 10 minutes and then closed the client application).

I've made full reinstall, all drivers are up to date, but no effect.

After I found this thread, I tried to reproduce the problem by running Linx test, but it always runs successfully without errors. Yesterday I disabled memory interleaving option in bios. I'll post the result here after some time. 

[Akorostel]

In addition to the previous post, I can report that after disabling "Memory interleaving" option in BIOS, both servers are working fine. There were no hangs or freezes during last month.

[aquilisdicio]

I am also having this issue with the same model and server 2008 R2.  I have tried disabling memory interleaving in the bios, without success.  Would appreciate a response from HP on this.

[ca336]

Hi Guys,

We purchased an ML110 G6 and Windows SBS 2011 essentials for a client of us. However, when trying to install the SBS essentials, we get the attached error.

The server has 12GBs RAM and 4 x 500GB HDs that 2 of each are mirror.

I will really appreciate any help on this.

[afamousguy]

I had the same problems and it took a new motherboard to correct the issue.  I am afraid to update the BIOS on the outside chance that it could be the cause.  I'm currently running version O27 (from 9/6/2010) with all default settings.

My advice to everyone with the same issues to keep pushing HP to swap the motherboard.  In my view, there is obviously a systemic problem with a run of these servers and despite the sheer amount of issues people are experiening it took a lot of time on the phone to get the problem resolved but it can be done if you press them.

[afamousguy]

@ca336 Are you installing off the DVD or did you copy the contents of the disc on a flash drive?  If you haven't tried it yet, track down the Microsoft tool which lets you dump the DVD to a flash drive.  I didn't have the same error when installing SBS 2011 Std but that fixed it in my case.

[Teun Pleunis]

I am having the same issue with 2x 4GB and 2x 2GB modules (12GB total). After one week of use the server started crashing randomly, usually within 24 hours and during load (backup etc).

I am able to reproduce the issue by running the program "memtest" (http://hcidesign.com/memtest/) 3 instances at once to test the full 12GB. Within the hour the server freezes again with no entry in the event log.

After replacing 2x 2GB modules the server was stable for one week, after that the crashes have returned.

After sending the logs HP is now going to replace the CPU. I will keep you posted.

[Teun Pleunis]

I tried disabling memory interleaving before HP got there to replace the CPU.

Did not help, still crashes.

HP replaced the CPU a week after. Crashes have now gone from once in 2 days to every 4 hours...situation got a lot worse...my customer is furious...

Will now try to get mainboard/cpu/memory replaced...

[Teun Pleunis]

HP has replaced motherboard and memory module (the 2 dimms I didn't yet replace). Ran memtest again, system crashed within 20 minutes.

I asked the HP engineer to remove the additional memory modules and run the server with 2x 4GB only. Memtest ran for 6 hours straight without any issues before I stopped it myself. Looking a lot better!

The server has runned stable through the weekend but I will have to see if it remains stable this week. Will keep you posted.

[Teun Pleunis]

Even with just 2x 4GB installed the server crashed after one week runtime.

After 2 months HP now believes it is a software related issue, not hardware. I do not agree.

My customer has requested to replace the entire server with a new one. HP does not agree. So to get rid of all the frustrations and putting expensive support hours down the drain I have bought a new ML110G7 server and am going to swap it myself. Hope this resolves the case.

Sadly this topic seems to have gone quite dead but I am still going to post updates to this issue.

[Teun Pleunis]

The new ML110G7 embedded b110i controller did not detect the RAID1 config I created and installed in the G6 server. Resulting in loss of all data.

To make matters worse Acronis Backup & Restore has compatibility issues with the b110i controller as well, detecting a RAID1 mirror as 2 separated drives...

After many headaches I have attached the disks to a P212/ZM controller (which Acronis supports), restored the backup using universal restore and, after booting, installed all correct drivers.

The new config has been running rock solid ever since.

In the end, what was wrong with the ML110G6 I still have here?

Was it hardware or just incompatibility with SBS2011? Don't know if I'll ever figure it out but it cost me in eccess of 40 hours in labour I did not get paid...

One of my lessions learnt; NEVER NEVER NEVER rely on the embedded b110 controller but buy P212 or better dedicated RAID controller so you can swap server systems without too much problems.

[colliema]

Peter,

i've been struggling with the random hang problem on my ML 110 G6 for quite a while. HP replaced the system board, but I still was having the issue, just less frequent. As it turns out, the server is VERY sensitive to memory configuration. I acquired the specific HP 1GB modules for this server, not a close part number, but the exact part number. The server is now rock solid for two months and counting. Hopefully this helps.

Mark

[CBROCK]

Hate to bump such an old thread, but I felt that I needed to chime in here.

My company purchased four of the ML110 G6's (PN- 597557-005) back in 2011 and all four of them exhibited the same freezing behavior. We eventually ruled out the memory, OS, array controllers, PSU, etc. and came to the conclusion that the mainboard is simply trash. Two were running Server 2008 R2, one running Server 2012, and one running VMware ESXi 5. All would randomly freeze up leaving nothing in the logs and needed to be rebooted (either physically by holding the power button or via iLO). Sometimes they'd remain up for weeks, other times it would happen twice in one day.

All of them had genuine HP memory as selected by the HP Memory Configurator tool. Nothing fancy here and nothing had been configured differently than any of our other 50+ HP ProLiant servers. We've even moved on to ML110 G7's and not had any issues with them whatsoever.

Point is, shame on HP for not extending the warranty on these affected products or not offering some sort of replacement path.

[jbwoodoo]

Hi,

I found this forum after struggling around with an old ML 110 G6, which I would like to use for training.  

After using easy install CD and running diagnostics from it, it showed up an memory ecc error ( Complete test ) , which confirms previous contributions in this thread.

I also disables mem interleaving and I also had change memory speed 800 Mhz. In this way I got rid of the memory ECC error.

My esxi 5 is now running like on other real servers.

By the way, I also did this funny firmware upgrades but it does not actually resolve anything.

Be aware and do not buy this model of servers (new or sec. hand), it looks like design fault.

Regards and thanks,

jbwoodoo