Abel's questions

I have been trying to dissect the information I find with googling, but have come to the conclusion that I cannot find with certainty the answer to the following questions:

Given you know the basics of your own desktop system, how do I know it supports NVMe drives?

• Does it require PCIe v2 or v3?
• Is a free x4 slot enough (if the spec of the drive says PCIe x4, for instance)?
• What does M.2 mean in this context (I have come to understand that it is an internal interface, and can connect with PCIe, mSATA or USB-3, depending on drive, is that correct?)
• Does the motherboard need to support NVMe (I think it does)? If it doesn't, can I still use the drive or is there an extension card?
• Does my BIOS need to support NVMe, or is "standard" UEFI BIOS enough?
• I read that the mean-time-to-failure is much higher (lower risk) with NVMe drives, is that because of NVMe, or because of the simple fact the more expensive drives have this and use better technology internally anyway?

I have come to understand from my previous question that any PCIe v3 device also fits in a PCIe v2 device, obviously with lower maximum speeds, is that also true for NVMe?

I know these are a bunch of questions and maybe I should've asked them one at a time, but I think they fit well together and may help others that struggle with understanding all the new techno buzz.

I am considering upgrading my workstation (almost a server) with a bootable PCIe drive, the 2TB version of the Samsung 960 Pro. But having a system with PCI v2, SATA II I am not sure if this makes sense as I am afraid of over-saturating the buses.

My main use for my system is data analysis and programming/compilation, where the data is oftentimes XML > 2GB and some of the systems I maintain presently take some 5-10 min to compile. During these tasks I can see that the CPU is doing next to nothing, so I am looking to decrease those times significantly by investing in SSD.

My system (it's from 2010, yes, it's old, but it still outperforms my high-end Asus laptop and I don't quite feel for replacing it yet):

• Dual Xeon X5680 @3.3GHz
• 48GB memory
• RAID-5 physical disks (measured at 150MB/s sequential), 3x 2TB
• 2TB Data disk (measured at ~250MB/s sequential)
• Nehalem chipset, Intel 5520 (Tylersburg 35D) with ICH10R I/O controller hub
• I have three free x16/x8/x4 PCI Express slots, which I think is capable of using that drive (is that a correct assumption?)

My current RAM disk is close to the expected speed of this drive (Read vs write seq. is 3,454MB/sec and 2,157MB/sec) with my current ramdisk (16GB of my 48GB is a RAM disk), I am worried the bus may be the limiting factor and not the raw speed of the drive:

I am particularly worried that my system supports PCIe v2.0 and that this drive requires PCIe v3.0 and this will limit maxing out this drive.

I am essentially looking for upgrading my system such that I get maximum speed gains with disk access. Probably any SSD would already improve things, but if I could get the best money can buy, would it make sense? Is there any sense at all in having such a fast drive in my setup, can use it to its maximum speed?

Small update: this discussion suggests that PCIe v3 is backward compatible with PCIe v2, so the slot should fit. The drive is x4, which I think means the max throughput is 500MB x 4 = 2GB/sec, which is below the maximum of this drive, but only for the (rare) sequential reads.

I found myself accidentally starting two large uploads, and since they are single BLOBs, I cannot easily stop/start them, potentially loosing much time. However, I prefer one of those processes to get the priority and give it, say, 80% of the available bandwidth.

With most file sharing tools it's relatively easy to limit or broaden the bandwidth. Is it also possible to do so for a running process? This is on Windows 7, but would equally be helpful on Windows 2008.

On Windows 7 and Windows 2008 SP2, how can I view who or what is using up my memory? In the task manager I can see the memory usage of user processes of me and other users, but there's always a gap between the sum of the memory of the individual processes and the actual total used memory.

I understand that several drivers need memory as well, for instance, I use a ramdisk, but it isn't visible in task manager, other than in the total memory graph. Also, it seems that a process can actually use more memory than the list of processes shows: when I killed explorer.exe, the drop in memory usage was about 1.5GB, while only 220MB was reported.

How can I query these "other" memory usages?

Not sure I use the right terms here, which is basically why I am asking (with the right terms I could google it for sure).

I have "one big server" with Windows 2008, with several virtual machines (linux, win7. win2k8) on it. The server is accessible with a public IP address. I would like to enable the following, and just don't know where to start:

• place websites on different virtual machines (i.e., VM1 hosts a.com and VM2 hosts b.com)
• sites should work with http and https (ssl) protocols
• allow remote desktop or similar RDP on each VM
• allow other protocols, like webdav, scp or ftp, to each machine

I kind of assume I need a VPN for the remote desktop access and I need some kind of host-based or host-based + protocol-based NAT (if that's what this is).

This seems like something that has been done a million times already, I just don't know where to start, what software to use or simply what to google for. Also, I'm afraid SSL wouldn't work this way because the hostname is not visible before the handshake.

I find the message hilarious and at first I thought it was some joke from an over-active programmer of the MS team, or a hoax. However, the message appears again and again, a couple of times a day:

NtpClient was unable to set a manual peer to use as a time source because of DNS resolution error on ''. NtpClient will try again in 3473457 minutes and double the reattempt interval thereafter. The error was: The requested name is valid, but no data of the requested type was found. (0x80072AFC)

Source: Time Service
Event ID: 134
Level: Warning

I suspect that my time servers are not configured correctly. Is that true and how can I fix it? But why such an odd message?

Note: I reported this on Technet, where you can find an explanation why the number is so odd (and consequently, two answers found that link too and used it in the answers ;) ).

For installation of a certificate I needed to install the root and intermediate certificates as well. This was recognized (after IIS reset) by Firefox and IIS at once, but not by IE, Opera or Network4All. After a physical restart it was recognized by all.

My question is: how can I install the root and intermediate certificates without restarting the entire server? (and, of less importance, come that Firefox recognized this at once, but others didn't)?

We have one site on IIS build in ASP.NET with DNN. Currently we try to configure network load balancing for two servers. The service hosting provider told us this would be enough, but we seem to have hit a brick wall:

• network load balancing only forwards requests from his public ip, it doesn't check the availability of the servers
• if a server goes down, 50% of the requests will see a "server not a available" until we manually remove the server from the load balancing configuration
• I don't see a way to configure this with two servers without this problem

With three-way load balancing, aka web farm, a single system checks which servers are available. If a server goes down, it will forward the request to the other server.

Is the above a correct assumption of the limitation of network load balancing, leading to classical web farm as only solution, or is there something the hosting provider can configure to prevent this behavior and have us stick with only two systems?

Several posts (like this one) seem to indicate that if you have a physical hard drive, you can "just" connect it with VMWare and it will be converted in a VMWare virtual machine.

I have a physical disk which is bootable (but not booted into!), and is accessible by Windows as drive H:. With all the options that VMWare offers to convert (a live system, a VHD image, etc) it doesn't list a way to simply pick up a physical drive and use it.

How can I convert this physical drive with a working OS to a VMWare image?

^{note: I also have a VHD backup (larger than 137GB), but not the VMC file, because I chose Full Backup from Vista; VMWare Workstation can only connect if the VHD is accompanied by a VMC file}

With my Dell Precision Workstation T7400 came a SAS controller that supports both striping and mirroring. Total nr of physical connections is 4. What I'd like to do is plugin four 2TB disks, select two of them for striping and the other two for the mirror of the first two.

I know, RAID-5 would be more suitable here, but that's not supported as far as I can tell. Is this a common scenario and should this be possible with my hardware and/or with a new controller card?

The morning, the following scary scenario happened:

1. I woke up my Windows system
2. Typed in my username and got an error (something like "could not load security xxx", but unsure of exact wording)
3. System auto-restarted after cliking OK
4. It didn't boot up anymore to the SSD with Windows 7 OS (I have another disk I can boot to, but that doesn't see the disk either).

Obviously, this happened right after I instantiated a backup procedure, which hasn't succeeded either. The BIOS can't find the drive when I connect to SATA. And it can't find the drive when I connect it to SAS. I have a Dell Workstation T7400, most recent BIOS (version A06), version of SAS Host Bus Adapter BIOS (HBA) is MPTBIOS 6.14.10.00 (2007.09.29) from LSI Logic Corp. Other findings:

• When connecting to SATA, the DELL Logo screen stays really long (5 minutes) and then at the end of POST it says that a drive is not found
• When connecting to SAS, the SAS HBA initializing phase takes long (2 minutes, against normally 15 seconds)
• When running Dell Diagnostics, it doesn't finish and gives the error Exception occurred in module MPCACHE.MDM file "IOAPICSP.ASM" line 1645.
• I contacted Dell. On their advice I tried different slots and different cables to no avail.
• I use an APIC battery power, spikes in the power are thus unlikely.

My conclusion so far: the disk is dead. I need this disk very badly because it contains the last few days of important development of which not all code was checked in the moment this happened. Are there any ways to recover dead SSD drives? The drive is a new X25-M G2 160GB model SSDSA2M160G2GC 2.5" in an extension bay and has been running without issues for 3 months on SAS.

When an NTFS volume is created, the first four entries of the MFT are copied and placed halfway the volume in the $MFTMirr file. According to Microsoft, since XP these files (metadata files) can be moved, but in practice I don't seem to be able to move them. Yet I need to reclaim a large contiguous free space block to be able to resize the disk.

I've tried several defragmentation programs, including the open source UltraDefrag, PerfectDisk 11 and Windows own defragmenter. But none seem to move this file, or the related $LogFile, $Bitmap etc.

What tool or command can I use to relocate these files from the middle of the drive to somewhere else?

I have a physical disk with two partitions, an old OS partition (with MBR) of 296GB and a RECOVERY partition of 2048MB (a default DELL recovery partition), both are not in use. I have successfully resized the OS partition from 296GB to 174GB with the following command:

DISKPART> shrink

which shrinks to the maximum shrinkable space. However, the partition only uses 56GB and I want to create a partition of at least under 128GB (to be able to create a VHD of it). When I shrink again, I receive the following error:

DISKPART> shrink desired=50000

Virtual Disk Service error:
The specified shrink size is too big and will cause the volume to be
smaller than the minimum volume size.

According to this post I should defrag, which I did using several tools (open source UltraDefrag seemed to do the best job). It seems that all files are on top of the disk, but how can I check that for sure? And if not, how can I manually defrag/move files, to claim the largest possible consecutive free disk space?

Or: what should I do to shrink the volume to its current used size plus a bit?

(Note: I tried to create a new VHD with the CREATE VDISK command pointing to the oversized VHD for the source-parameter, but here too, the 128GB/137GiB limit prevents success, it seems. Mounting the VHD and shrinking it had exactly the same effect as doing it on the original physical disk, which is why I now try to first limit the physical partition size.)

If you have many sub-domain names like xxx.example.com, xyz.example.com etc, you can solve these through server-side scripting or other means by using a wildcard A record for *.example.com in your DNS.

How can I determine whether a wildcard domain is configured for any given domain name? Using http://network-tools.com gives a lot of information, but doesn't reveal this. If I need to use commandline tools: I use Windows. One such example that uses a wildcard domain DNS, I think, is blogspot.com.

We recently upgraded to Windows 7 64 bit (workstations) and Windows 2008 64 bit (servers). To prevent loosing configuration information, and complex settings, we created a multi-boot system when upgrade was not option (i.e., when original installation was 32 bit, you cannot upgrade to 64 bit).

Simply mounting the volumes is not what I'm after. I've got some backups as ISO files, these are inherently read-only, and need to be converted / unpacked to be able to run them as virtual machine. Others are on bootable partitions.

Given a physical and bootable Windows Vista, XP, 2003 partition, what virtual machine software can I use to access it if that partition is on the same physical machine I currently run, preferably without conversion or copying.

While troubleshooting a Samba vs Windows Network issue, I noticed that Windows' Explorer remembers login credentials of remote shares, even if you ask it not to. For instance, after accessing a share using \\servername\sharename plus entering username/password and then closing Windows Explorer, adding the same share as a network drive gives the following message, regardless whether the username is the same or not:

The network folder specified is currently mapped using a different user name and password.

To connect using a different user name and password, first disconnect any existing mappings to this network share.

Using NET USE does not show the share. After restarting the computer, I have no problems accessing the share using different credentials. But restarting just for testing other credentials is annoying, esp. while troubleshooting. How can I purge this cache, using Windows Vista?

^{Note: using nbtstat -R[R], ipconfig /renew, killing explorer.exe or disabling / re-enabling the network card didn't help.}

If you configure Exchange Server 2007 for auto-reply on a public address, and the reply is sent to an address that also has auto-reply or out-of-office-auto-reply, then Exchange Server will receive a new message, same from-address, and will again send an auto-reply. This can go on forever and can potentially lead to a DoS situation.

How can I prevent multiple sending of auto-reply to the same address, preferably in a given timeframe (to prevent legitimate multiple mails to be treated incorrectly), using Exchange Server 2007?

Our XS hosting provider says it cannot be done, but that strikes me as odd.

We have severe performance issues on an ESX client, but from where I am I cannot directly view the server configuration, yet I want to find out whether they set it up correctly.

Running CPU-Z shows me two processors with each one core, while I have Yorkfield processors and hence expect four cores to show up. Is this the way virtual CPUs show in CPU-Z?

How do I determine, from the client, how ESX has it configured? I understand that ESX can have HT enabled. My system uses SMP heavily, so it should benefit. From what I gather, the ESX is configured for two vCPUs. Is my reading of the data correct?

EDIT: there are basically two causes for the choking CPU. First, the application(s) aren't perfect, we work on that. Second, we don't have enough CPU power. The image comes from a similar native machine and is now on a virtual machine. Performance should be similar, but isn't. Normally we could deal with 500+ sessions per native system. On this system, choking starts around 280 sessions. I know it's a bit short-sighted to blame the CPU config, but after two days intensive monitoring and profiling it all really starts to point in that direction (memory, disk IO, db connection and network are performing fine).

For a while now, I'm trying hosting providers (ASP.NET at least). What all tried providers have in common so far: a more-a-less silly password limitation policy that ridicules security.

Correct me if I'm wrong, please, but I came across the following, asked the providers how I could trust their security if they disallow strong passwords and they all answered "your data is safe with us, we're a secure provider, look at our customers etc".

Between the hosting providers I tested, I found among others:

• Password length <=8, only letters / digits
• Password length <=8, only letters / digits + list of some other chars
• Password length <=8, unclear what's allowed, server error on some chars
• Password length <= 12
• Inside Plesk: password length restrictions for databases
• Passwords send in plain text, or shown on Plesk or other admin tools
• Invalid, outdated or low-keyed SSL certificates
• etc

Am I right in saying that this makes me (or the sites I host) vulnerable?

This seems common. Does this only happen on the low side of the market? Am I right that only VPS or colocation are secure bets? Do you know of any godaddy-level priced hosting providers that do not pose limitations or should I seriously go up?