Servers

Servers power many organisations and are even making their way into homes (even if it is a simple NAS box). With more RAM then you could even imagine (I have to admit I’m a RAM addict) and processors that make the latest and greatest gaming PC look minuscule in comparison it is these mighty machines that make the world go round.

As some readers may know I currently work in Higher Education and while all of the business data is trivial to backup providing any level of backup service to students and academics is significantly harder. The challenges faced include the myriad of Operating Systems in use (Windows/OSX/Linux), the fact that the devices being backed up are inherently ‘untrusted’ (i.e. owned by the individual) and that they are often on networks (be it eduroam/public/home) that have no direct connectivity back to the internal trusted network.

Most enterprise class backup systems just aren’t suited to this kind of environment in that they cannot be securely published through a firewall or have exorbitant licencing costs for the number of devices to be protected (a few file servers vs 500+ student owned laptops).

One solution to this issue cropped up at a recent trade show where Synology were demonstrating their Synology DiskStation Manager NAS software which set itself apart from the traditional enterprise backup solutions with…

  • Support up to 16,000 users on high end models (and 2048 on the kind of model that we would consider using) with no extra licencing costs, users can have storage quotas set either by group or per user
  • Secure remote access (simply publish a single port which can be protected by HTTPS for encryption in transit)
  • Home grown backup clients for modern versions of Windows, OSX/macOS and Linux
  • On the point of OSX/macOS the backup client for Synology does not rely on Time Machine and so overcomes the issues associated with having to be on the same network as your backup device
  • Home grown Btrfs file system which auto detects (and fixes) corrupted files through metadata along with extensive snapshot support
  • Up to 32 recovery points and real-time file protection (when connected to the DiskStation)

So time for some screenshots! Below we have the initial setup of the Disk Station Manager and the installation of the client on a Windows PC.

Then restoring a file that has been deleted on the Windows PC; note that you can restore either individual files or entire folders to a point in time.

The same but for OSX…

So that’s all of the good, the only downside we have found thus far is while shared drives can be protected with encryption it is not possible to protect each individual home area (per user) with a unique encryption key thus opening up issues with data privacy. However, if you consider the following scenario…

  • A business needs to provide backup to remote workers
    • Those remote workers do not connect to the trusted network often
      • Perhaps they don’t like VPNs/DirectAccess (and so rules out using Offline Files)
    • and those remote workers do not use a commercial ‘cloud’ service to protect their data with
      • Perhaps trusting a 3rd party to host the data is not an option
    • The remote workers use OSX/macOS

…then using a Synology DiskStation should be a serious consideration for that business.

When working with a lab full of HPE ProCurve/Aruba switches (or you just want to know who is who in a stack of switches) the chassislocate CLI command comes in really handy by either blinking or holding solid the blue locator light. See the screenshots below for a little more info.

It’s that magical time of the year where…new network switches arrive! Given that the new Aruba branding has taken full control of what was ProCurve I thought I should post some photos of the new paintwork. Happy to say the colour black isn’t half bad!

Included in the images are

  • J9729A 2920-48G-PoE+
  • J9728A 2920-48G
  • J9731A 2920 2-Port 10Gbe SFP+ Module
  • J9733A 2920 2-Port Stacking Module
  • J9734A 2920 0.5m Stacking Cable

1

Picture 1 of 3

Error 8024400E appears on the clients...

After recently deploying a Windows Server 2012 R2 WSUS server (afraid we couldn’t wait much longer for 2016 (which is now out by the way!)) we started seeing Error 8024400E on our clients and servers (from 2008R2/Windows 7 to Server 2012 R2/Windows 10).

As it transpires Microsoft published KB3159706 for the WSUS server which adds some new features to be able to manage Windows 10 updates and thus requires some manual post installation steps which can be found at this link here – https://support.microsoft.com/en-gb/kb/3159706.

The steps only took a few minutes to go through so it was a pretty easy fix in the end.

In this guide I’ll show a ‘working’ method to upgrade from MySQL Server 5.1 to 5.5 on a Windows Server 2008 R2 machine. In this case MySQL Server 5.1 was installed by the Microsoft Web Platform Installer some time ago, however this version doesn’t deliver very good performance (and its horribly outdated!) so it was time to get it replaced.

The good chaps at MySQL do offer a guide on how to do this on their website – https://dev.mysql.com/doc/refman/5.5/en/windows-upgrading.html however I hope you will find this one will serve you better.

A few points to note before we continue….

  • This guide assumes you are using a default install of MySQL Server 5.1 (as delivered by the Microsoft Web Platform Installer)
  • This guide only looks at a ‘simple’ server deployment i.e. no clustering or other funky features are in use
  • It is recommended to jump only one version at a time when going from MySQL version to version e.g. 5.1 to 5.5 (there were no 5.2, 5.3 or 5.4 versions) and then 5.5 to 5.6 and 5.6 to 5.7…
  • You can download previous versions of MySQL Server from this link – http://dev.mysql.com/downloads/mysql/ just click on the ‘Looking for previous GA versions?’ link
  • Be sure to take a backup of your databases before you attempt this guide on a production machine (see https://dev.mysql.com/doc/refman/5.5/en/backup-methods.html for some ideas on how to do this)
  • If possible run through the upgrade process a few times on a testing machine (one that you can break and no one will notice) first

So on with the guide!

Final point to note….this guide only works with migrations to 5.5 from 5.1, if going to further versions you will be missing some tables which will in turn generate error messages during the upgrade process. More information on that here http://dba.stackexchange.com/questions/54608/innodb-error-table-mysql-innodb-table-stats-not-found-after-upgrade-to-mys.

Work is coming along nicely with the Server Room, we’ve now removed the last Cisco switch from our infrastructure and the HP 5400R series switch is deployed replacing the 2530 that was in its place; over time we’ll be bringing more fibre from our edge switches into this room as well hence the number of SFP+ ports on the 5400R. The entire front of the cabinet is now populated with hardware or a blanking panel as well (panels available from Comms Express) to keep things looking tidy. I wish there were a little more that I could do with the cables coming into the 5400R however with a very narrow rack there’s not much that can be done.

Some interesting things have come out of both Rucks and PaloAlto recently in that they offer Hyper-V compatible VMs for their services which could free up a further 3U of space and remove a further 4-6 cables out of the picture.

Looking for some fun ways to get more out of your your Smart Card deployment? If so have you tried……?

  1. Use Smart Cards to login to your Servers via Remote Desktop
  2. Use Smart Cards with the PowerShell Get-Credential Commandlet
  3. Use Smart Cards with your Firewall for single sign on
  4. Use Smart Cards to login to IIS Web Applications (just a box to tick and a radio option to select)
  5. Store multiple identities on your Smart Card and assign different (and perhaps more complex) PINs to the identities

Have a look at the screen shots below for some more details…

Check DCsOne of those monthly jobs that every SysAdmin will come across is good old Patch Tuesday; to help make Patch Tuesday a little more fun after all of the servers have been updated I use Hyper-V Replica (run by a PowerShell script) to shutdown each Virtual Machine and move it onto another host (ticks the box for the machine reboot component of Windows Updates and also tests our DR solution in one hit!).
However as both of my DCs are Virtual Machines I want to make sure that at least one DC is up at all times, to do that I have built a little PowerShell function (see below to download it within a zip file!) that is run before every migration to ensure that both DCs are up and running (along with the Network Policy Server service which is used to authenticate clients on the network (and so is very important!!)) before any migration happens.

Hopefully this will help someone someday!

  See if DCs are up (905 bytes, 344 hits)

A silly gotcha more than anything else…. after recently updating my WSUS server to use SSL (to allow publishing through the firewall) I noticed my clients that were deployed with MDT (Microsoft Deployment Toolkit) were not installing updates as part of the Task Sequence; indeed the message log at the end indicated that the updates could not be downloaded as there was no connectivity to the WSUS server.

Lone behold I had updated the path to be https:// (against http://) in the Group Policies that pointed the clients at the WSUS server but not in the Deployment Share properties in MDT. So let the lesson be learnt… be sure to make the URL change in MDT as well as in Group Policy.

1While working on my most recent Hyper-V Replica PowerShell script when attempting to reverse replication from a source Hyper-V host to the a target host using Certificate authentication I was getting the error message…

Hyper-V failed to establish a connection with the Replica server ‘<target hostname>’ on port ‘443’. Error: The connection with the server was terminated abnormally (0x00002EFE).

As it turns out I had recently deleted and created a new certificate for the target host and as such there was no certificate listed in Hyper-V Settings > Replication Configuration. The fix was to set the replacement certificate in the box provided. See the screenshots below for a little more…