飞来飞去的日子：他们来自他方

成了Trainer 后，看东西的角度也不一样了。以前在Tech Support部门的时候，很痛恨业务部，觉得他们为了得到可观的数额，把产品‘乱乱卖’；直到客户发现产品并不是像业务说的那么‘神’，就会致电到技术支援部来大骂一番。

在公司将近8年了，渐渐发现这是因为现有的政策不够完整，导致技术支援和业务从来没有机会互相了解彼此的工作岗位；两个部门只有在客户有问题的时候才会有所接触，那时候已经没有让他们了解的时间，换来的只是一些不必要的摩擦！

日复一日，7年后的今天，技术支援和业务两个部门还是在犯同一个错误！这究竟是谁的错？

飞了那么多地方，也遇见过形形色色的人，结论：每天等待5点钟下班回酒店好好歇息！
感谢JBL OnBeat 迷你喇叭陪我渡过无数孤独的夜晚...

他们来自斐济（Fiji）、巴基斯坦、蒙古、新加坡和马来西亚....

这样的工作态度还可以生存？

我又扛了编制时间表的责任，早知道会有问题出现，只不过没想到问题会出现得那么早而已。对于经理的决定我也只能觉得很无奈，我想他也有苦衷，也只能支持他了。

几天前，刚拿了这个责任，我尝试编排一个最没有问题的时间表。思前想后，尽了力还是只能将整体处于平衡点，还是会对其中几个同事有点不公平。但是，这不是常有的事，也觉得大家出来工作的，偶尔有这种状况也是在所难免，就直接将这个时间表发给了每个人。

其中一个觉得对他不公平的，是因为刚从晚班出来，就直接要在星期六（OT）上班。我也明白，不过这是跟着顺序而编排，而不是我故意刁难他；再者，他的晚班在星期四早上已经结束，至少还有整个星期五可以歇息，况且星期六并没有什么客户打电话进来，也可以轻轻松松的上班啊！我先帮他问了好几个同事，也尝试看看在那个时间可以怎么调整，不过，真的没人顶替了。我也硬着头皮跟他说记得来上班了...他，回了我的SMS如下：

“...In that case u swap with me la, i rather sleep at home...”

说真的，看了这封短讯，我真的很气！这难道是他父亲的公司？可以选择上班与否？刚读完那封短讯，另一封短讯又到了我的手机：

“...i might still help if someone were to asked me earlier before auto-assigned to me directly.At least I have option you know...”

如果我是他老板的话，这样的工作态度，在这个Team需要帮忙的时候却显出了这样的态度，这根本不是领导者的素质，也没有Team Work可言，日后肯定不会重用他。

算了，本来想要告诉他可以选择星期六不必上班，然后我来顶替；一番深呼吸后，还是和经理讨论了。最后决定将那一时段的Shift废掉一次，他，可开心了...

没关系，我终于认识他了...

一天一天看着他长大，虽然时常为了教好他而吼他，但是心里确实很疼爱他。不认识我的，肯定觉得我很没有爱心，没关系，只要我知道我有就好。因为一个人根本不可能为了他人的眼光而活在这个世界上...:-)

Power Consumption for Dell PowerConnect 6224P

Being a Server/Storage support for three years, I can't believe the below enquiry was first time delivered to my ear:

"..we already confirmed the order for Dell PowerConnect 6224P, and it should arrive tomorrow. Before that, we would like to know the power consumption for this switch, so that we can plan for the power input source..."

It sounds nothing wrong, but the documentation is killing me again because it just leave me with this lines:

"The 6200 series switches have one internal power supply which requires standard AC"

Surprisingly, this Singaporean doesn't know (or maybe pretend do not know) what is standard AC - Alternating Current for their own country. After research for almost one hour, I managed to get help from Google about the whole complete list of Singapore Standards -

SS 307 : 1996 -> Applies to portable cable reels for alternating current (a.c.) only, having a rated voltage not exceeding 250V, rated current not exceeding 13A and provided with a non-detachable flexible cable or cord, intended for domestic and similar purposes, either indoors or outdoors. Does not apply to cable reeling devices incorporated in appliances.

Luckily this typical Singaporean was satisfied with the provided information.

The simplest question always will become the most difficult one to answer.

one of my favorite sunrise taken from Penang Jetty...Guess what, this is also taken with my D60! :-)

What an "experienced" Server Administrator

I dare not describe myself as a meticulous type of person, but I do hear my customer very carefully, especially when I got a call from a Malay speaker (due to my 'bahasa' isn't good enough).

"Can you help me to disconnect the power cable from the back of the server?"

"Ok, let me go to the back of the server..", said the Malay lady.

After 3-4 minutes, the Malay lady came back and said

"I tak nampak tu power on-off button dari back lah, you sure ni dekat server punya belakang ke?", surprisingly this lady start to doubts on me.

My intuitive sense told me this so call Server Administrator doesn't have any technical knowledge on server and I have to treat her like a primary school girl....

Lastly, I manage to end this call with my glib instruction to guide her through out the troubleshooting, and I'm sure my blood pressure increase a lot while I'm talking to her. Cheers for this company to have such 'experienced' server administrator. -.-"

Penang Kek Lok Li's Pagoda. It has been there for decades, and I just make my first visit to climb up this pagoda last Chinese New Year. What about you?

Dell Software RAID Controller S100/S300 not compatible in Linux OS

Recently, Dell offered new software RAID card, with code name: S100/S300. Beware that currently this RAID card just compatible to Windows OS, and yet to see for future improvement that can work with Linux OS as well.

Specification	PERC S100 controller	PERC S300 controller
SAS technology	No	Yes
SATA technology	Yes	Yes
eSATA technology	Yes	No
SSD technology	Yes	No
Support for x8 PCI-E Host Interface	No	Yes
I/O Controller	Intel™ ICH10R or Intel™ Ibex Peak chipsets	Dell 3Gb/s SAS Adapter
Communication to the system	Integrated	PCI-E lanes
Communication to end devices	SATA links	SAS/SATA links
SAS connectors	No	Two 4-port connectors on all systems
SATA connectors	Discrete on the motherboard	Two 4-port connectors on all systems
Lead-free	Yes	Yes
Supported operating systems	Microsoft Windows Server 2003 Family, Microsoft Windows Server 2008 Family, Microsoft Windows Server 2008 R2
Dell-compliant SATA compatibility	Yes	Yes
Dell-compliant SAS compatibility	No	Yes
Dell-supported direct-connected end devices	Dell-compliant physical disks	Dell-compliant physical disks
SMART error support through management applications	No	No
Backplane supported systems	No	Yes
Software-based RAID	Volume, RAID 1, RAID 0, RAID 5, RAID 10	Volume, RAID 1, RAID 0, RAID 5, RAID 10
Maximum number of virtual disks	8	8
Support for internal tape drive	Yes	No
Support for global hot spare	Yes	Yes
Maximum number of hot spares	Varies (by the number of free disks in the system)	Varies (by the number of free disks in the system)

Reference: http://support.dell.com/support/edocs/storage/Storlink/PERC_S100_S300/en/HTML/a5_sectd.htm#wp1201881

Dell Remote Access Card (DRAC5) not able to send alerts to IT Assistant

If you wish to configure Dell Remote Access Card (DRAC5) to send alert to IT Assistant, there is one option that you need to take note in order to make it works.

1. Log into Remote Access Web Interface, click [System], then go to [Alert Management] tab.

2. Check the [Enable Platform Event Filter alerts] checkbox. (Default Setting is unchecked).

3. Then, perform the hardware reset of the DRAC as below:

i. Click the Start button, and then click Run.The Run window appears.

ii. In the Open text box, type Command, and then click OK.
The Command Prompt screen appears.

At the C:\ prompt, type racadm racreset, and then press the key.

*before you're able to perform this configuration, please ensure your SNMP Services is in working condition

Dell PowerEdge: T310 failed to install Windows Small Business Server 2008

Last week, one customer from Singapore called up and blame Dell Server again. According to this customer, he said Dell had posted up a false info over the internet. This is because he failed to install Windows Small Business Server 2008 to Dell PowerEdge T310, and documentation written that this OS is actually supported on this platform.

This customer had done a lot of research before ring me up, and wanted to send me the T310 documentation link to prove he is right.

I knew he was right, so I check his server's hardware specification, found that he only having 2GB RAM for this server. At first I really didn't know what's wrong with this, inform customer that I will need to call 1 hour later, and I google for more info about this.

I'm lucky enough to remind myself that one of my colleague share this info before. Then I read through the article, confirm this is the answer that I can share with this demanding customer.

Cause:

He didn't meet one of the minimum requirement of having at least 4GB to complete the OS installation.

Plan to switch to Windows Small Business Server 2008 Standard/Premium? Please refers to below minimum requirements.

For more information, please visit Microsoft site as below:

http://www.microsoft.com/sbs/en/us/system-requirements.aspx

Dell PowerEdge: Why R410 only 4 DIMM slots are usable?

Many customers didn't really know what to ask before confirm the order, but rather accept what Sales person had suggest to them.

No one is wrong in this situation but I just share it out for those who like to research before buy the product.

I would like to share about Dell PowerEdge R410's Memory Configuration.

Two days ago, I received one inquiry from a customer.

Problem:

He bought R410 with 2x 2GB DIMM. After that, he purchase 4x 2GB DIMM from another sales. After he connect all the DIMMs, found that system BIOS only detect 8GB, in other words, only 4 slots are working.

Then he is blaming me why the new server (just arrive for less than 7 days) already having hardware issue. I check the correct memory population, and take a look at the hardware component that he order, then I know the answer.

Cause:

This is due to customer only purchase with 1x Processor. This is why only 4 slots are working.

Workaround:

I suggest him to keep the DIMMs for standby purpose, or consider to add in another processor to get all slots detected.

Dell PowerEdge: R610 / R710 only detect my 1333MHZ DDR3 RAM as 1067MHZ in System BIOS

Nowadays server already using DDR3 Memory in most of the platform regardless it is Dell, IBM or HP. For this new technology, it have two types: Registered (RDIMM) & Unbuffered (UDIMM) ECC DDR3.

The primary differences between UDIMMs and RDIMMs are:

• RDIMMs have a register used as a pass through for address and command signals. This allows servers to support more RDIMMs per channel than UDIMMs.
• RDIMMs consume more power than UDIMMs but are available in larger capacities and higher ranks.
• RDIMMs also provide additional data protection and reliability features.

(Reference: Memory Selection Guidelines for High Performance Computing with Dell™ PowerEdge™ 11G Servers)

I'm not going to go into detail of the difference of RDIMM & UDIMM, but I would like to share the situation below:

Customer contact me and request my explanation about why new server's motherboard is faulty. Then he share me the screen from System BIOS, that all 1333Mhz Memory are detected as 1067Mhz only.

I checked the Intel Processor that he ordered (Intel Xeon E5520 2.26Ghz), then I get the answer.

This is nothing about the motherboard, or the design of Dell platform, but this is due to the Limitation of Intel Xeon Processor.

Whitepaper (Page.11)

From the chat above, it is clearly stated that only X5550 and above able to support/detect 1333Mhz of DDR3 RAM (but still depends on processor model, DIMM speed, and DIMM population).

I hope my sharing over here can avoid further misunderstanding towards Dell platform. :-)

Dell PowerEdge: Intel Nehalem Processor cause Blue Screen: STOP 0x00000101

When 11th Gen Dell PowerEdge Server being introduce to the market, it face a lot of challenge from other manufacturer that make customer have concern about part quality and server stability.

One of that challenge is Intel Nehalem series processor can cause blue screen "STOP 0x00000101 - Clock_watchdog_timeout" when you activate Hyper-V from Windows Server 2008 R2.

Two months back, I have customer called in not only scolding me like hell but would like to know why Dell server always cause blue screen, and not stable as what its described from Sales person.

You can see the below information from Event Viewer:

Problem signature:

Problem Event Name: BlueScreen

OS Version: 6.1.7600.2.0.0.274.10

Locale ID: 1033

Additional information about the problem:

BCCode: 101

BCP1: 000000000000000D

BCP2: 0000000000000000

BCP3: FFFFF88001F46180

BCP4: 0000000000000004

OS Version: 6_1_7600

Service Pack: 0_0

Product: 274_3

Files that help describe the problem:

C:\Windows\Minidump\021110-36363-01.dmp

C:\Users\administrator.PANDALOG\AppData\Local\Temp\2\WER-1299472-0.sysdata.xml

I have no choice but get the stop code error from customer, and start my research from Microsoft Knowledge Base(KB), as well as using WinDebugger to analyze the Dump file.

Finally, I found one KB from Microsoft that explain the same STOP error, but it pointed to Intel Nehalem Processor's limitation. It simply explained this is due to serious interrupts generated from Intel code-named Nehalem Processors. For more information, can refers to below Intel document:

http://www.intel.com/assets/pdf/specupdate/321324.pdf

Microsoft KB:

http://support.microsoft.com/kb/975530

"Don't judge a person by his appearance."

Dell Server: How to capture hardware log if using VMWare or Open Linux?

Good news to all Dell Customer who using VMWare or other Linux OS which not compatible with Dell OpenManage Server Administrator (OMSA).

Now, you can use Dell LiveCD to boot up during POST, and access to Embedded Server Management (ESM) to gather the hardware log.

LiveCD ISO Download:

http://linux.dell.com/files/openmanage-contributions/omsa-601-live/omsa601uk.iso

How to use?

1. It is a bootable CD, after you burn the ISO into CD, just boot the CD or change the boot sequence in BIOS settings.

2. Log in User Name= root ; Password = dell

3. After logged in, it will display the current IP Address, copy down as you may need it later for remote access connection.

4. Next, just type 'startx' to access the GUI

5. Then you'll be able to see one browser icon with the name "OMSA".

6. Double click to open, and you'll be able to see the hardware status as the normal OMSA in Windows environment.

How to Create RAID0 with single HDD by using Dell SAS5iR / SAS6iR?

If you ever go into SAS5iR / SAS6iR configuration BIOS, you will notice that it require at least 2x Hard Disks to configure RAID0. I have to say this is just as per design, and it is not refers to bug or limitation.

Some of you may wondering if you're using Windows Small Business Server 2003/2008, which have the limitation of having 2TB for C:\ , and yet you don't want to convert it into Dynamic Disk; or you totally don't need that huge space for your C:\, and the server already in front of you. Stop to blame the design or the Sales person of not inform you about this, but refers to below workaround.

Workaround (I):

1. Go to SAS5iR /SAS6iR Configuration BIOS which is Ctrl +C, delete all RAID0 or RAID1 configuration.

2. Exit the BIOS, then proceed the OS installation through:

11th Gen Server:

~ Unified Server Configurator (USC) or Installation DVD (Dell System Tools & Documentation)

8th - 9th Gen Server:

~ for this generation, do not support USC, so we need to depend on Installation CD/DVD which can ease the installation process of end user.

3. However, with treated this card as a 'pass-through' SAS interface controller card, you will notice one of the HDD LED light (status LED) will not light up.

Dell did explain that in their Knowledge base:

Drive Status LED Functionality When Attached to SAS5iR and SAS6iR

Explanation:

When hard disk drives are attached to a SAS5/iR or SAS6/iR as a single disk (not part of a RAID0 or RAID1) the status Light Emitting Diode (LED) will not illuminate. The LED is not a representation of power to the drive. It is a representation of the drive status or availability within an array. Since a single drive Redundant Array of Independent Disks0 (RAID0) cannot be created on a SAS5/iR or SAS6/iR, the status light will remain off for single drive usage.

Workaround (II):

1. Check with Technical Support or Sales whether the Model of the server that you own is support PERC series of SAS RAID card or not.

2. If so, you may consider change the card to PERC6i (for 11th Gen) or PERC5i (for 9th & 10th Gen).

3. PERC series able to create RAID0 with single HDD.

*However, the price of PERC series is much more expensive than SAS5iR/ SAS6iR.

Dell Server: PowerEdge R710 with Hyper-V will hang/freeze

Recently I encounter a case with below situation:

Product: Dell PowerEdge R710

RAM: 12GB ECC DIMM

Processor: 2x Quad Core Intel Processor

OS: Windows Server 2008 Enterprise R2 x64

Others: Hyper-V activated

Problem:

Customer mentioned that he have total 5 VMs running in this server. 4 of it will running all the time. He allocate 8.5GB for 5 VMs, and after 2-3 days, server will hang. He monitor this issue by his own, and found that Memory usage will reach 100% before it hang.

He submit this issue to me, and cause me study a lot of whitepaper, forum from Microsoft site to look for more information.

VMs Resource Allocation:

1. VM1 Memory=512MB HDD Space=80GB

2. VM2 Memory=2GB HDD Space=200GB

2. VM3 Memory=4GB HDD Space=300GB

2. VM4 Memory=2GB HDD Space=150GB

In a nutshell, he left 3.5GB RAM for Host which is totally insufficient to run in a optimal environment.

Maybe you'll ask: "Where is the 5th VMs?"

I ask the same to customer, but he told me the 5th VM in OFF mode all the time, will use when its needed. -.-"

Research:

After 1 day of studies, research, I manage to obtain the below info to customer. To configure Hyper-V with Optimal state, we have to follow the below formula:

To calculate the host RAM = (VMRamRequirementMB x #ofVMs) + (#of VMs x 32MBVMOverhead ) + (512MBHost)

To calculate VMRamRequirementMB, I found 1 statement from one of the Microsoft White Paper:

...the root partition must have sufficient memory (leave at least 512 MB available) to provide services such as I/O virtualization, snapshot, and management to support the child partitions.

A good standard for the memory overhead of each VM is 32 MB for the first 1 GB of virtual RAM plus another 8 MB for each additional GB of virtual RAM. This should be factored in the calculations of how many VMs to host on a physical server. The memory overhead varies depending on the actual load and amount of memory that is assigned to each VM.

Reference:

http://download.microsoft.com/download/7/E/7/7E7662CF-CBEA-470B-A97E-CE7CE0D98DC2/Perf-tun-srv-R2.docx

(Search for “Correct Memory Sizing”)

Therefore, I manage to come out the optimal resources allocation with above-mentioned formula:

1. VM1 = (1 x 32MB) + (79 x 8MB) = 664MB

2. VM1 = (1 x 32MB) + (198 x 8MB) = 1616MB

3. VM1 = (1 x 32MB) + (299 x 8MB) = 2424MB

4. VM1 = (1 x 32MB) + (149 x 8MB) = 1224MB

Total RAM for VMs = 5928MB

= 5.8GB

Host RAM = (5928MB) + (4 x 32MB) + 512MB

= 6.41GB

Result:

According to calculation, obviously server (host) doesn't have sufficient Memory because it need at least 6.4GB, and VMs already occupied 5.8GB (maybe more as I didn't count in the 5th VM that customer turn on when it is needed).

Temporary workout:

1. Reserved at least 6.4GB for Host. This can only avoid the Host from hang/freeze issue, but it may cause VMs couldn't run in Optimal state.

2. Reduce the size of HDD Space for the VMs.

Permanent solution:

1. Add in more RAMs as Widows Server 2008 Enterprise R2 able to support up to 32GB memory space.

2. Move at least 1 or 2 VMs to other Host to reduce the loads in this server.