0

I just started a role as the sole "IT guy" in a company of over a hundred people. This is a company that has had a troubled history with IT, managing to go through 6 people in as many years. During that time each person has barely documented anything, has had different ways of doing things, and all systems have managed to grow organically as more items have been tacked on without any effort to build a solid foundation.

Plus side : The company has realized that trying to put everything on one person has not been working. I will be the first IT hire in what will be a reasonably sized team. Additionally we will be outsourcing a general audit of our systems but this audit will be solely focused on our client-facing production systems and not the network as a whole.

Some interesting facts :

  • There are a variety of storage systems in place from different vendors. There is one storage system where no credentials have ever been saved and it is starting to fail.
  • The network as a whole doesn't make much sense to me...6 domain controllers to service one location/domain. Multiple VLANs that seem to have no rhyme or reason. There are multiple satellite offices off on their own domain that have been completely neglected by the last few of my predecessors.
  • No backups appear to be taking place
  • Office365 is in place to service email/lync/sharepoint but does not appear to be synced with active directory
  • Speaking of active directory. . . it is a complete mess. Mostly unorganized, some ancient OUs that have actually been causing problems, almost everyone is a domain admin.
  • No ticketing system or asset tracking has historically been used, ever.
  • Across the board there are communication issues machine to machine. There is a cisco ASA where almost all traffic in-network seems to be routed through and it is very barely configured.

These are just what I've noticed the first day. I've also been told that we probably have a serious critical failure once a quarter.

This amounts to a pretty big challenge and I'm pretty stumped at where to begin, so I figured I'd ask everyone here. . . where should I start?

Improve this question
6
  • 3
    Leave the job to someone who does not ahve to ask this. Seriously - this is a HUGH field and way too large to answer here. This is the case for a very senior person who may take a year or two to migrate all this to a sane setup. – TomTom Apr 15 '15 at 18:05
  • 6
    Start by verifying you have fully functioning backup systems that cover all your data/configuration. – Zoredache Apr 15 '15 at 18:27
  • 1
    Yeah. FIrst thing likely is backup. WIthout it you can not change anything anyway and run risk of catastrophic failure. – TomTom Apr 15 '15 at 18:30
  • 2
    @Zoredache: 1000x times this. Check it, test it. No one cares it's not your mess if something breaks and you don't have a working backup. – Sven Apr 15 '15 at 18:31
  • 3
    Don't touch that horror until you have working backups. – user186340 Apr 15 '15 at 19:29
5

You mention client-facing systems. This to me means your first priority is uptime on the systems that make you money. Networking and storage sound like #1 and #2. (Edit:) Oh, and resiliency. Get your backups checked out, in-line and tested. Make sure you can restore them!

Prioritize internal network and AD next, you need a solid foundation. Based on your info, I would say a migration/domain collapse is in order. I cannot imagine a reason to have multiple AD domains for such a small org.

Get a project manager. You will probably have multiple concurrent efforts to get his cleaned up. Have a pro coordinate work.

Don't be afraid to outsource some of this to professional services. Scope it as tightly as possible, and if a vendor impresses you, look at expanding engagement.

Improve this answer
3

As stated above, this is a question with a multitude of very long answer.

So, I'll try to keep very short:

  1. backup: with no backup, if something breaks you really have a change of going out of business
  2. storage: check that your storage servers (and their disks) are in good healt
  3. networking: a careful policy audit should not take too much time (about 1-2 weeks)
  4. virtualization: start migrating your physical servers to virtual machines
  5. after that, you can start reorganizing your domain at a software/application level. In this case, make ample use of VM snapshots
Improve this answer
1

I'd search for the help of a project manager on a scenario like this. From my experience, tackling network (infrastructure and logic), backup and security first is a better way to pavement the changes. Also, document every aspect of the current state the whole thing runs before making any change to the environment (photos are great depending on the case).

I had a role similar to yours in the past. It was a gov't company running without a budget at all. The 150+ workstations and 10 servers ran into non-managed 10/100 switches and I've seen almost every problem: From Pentiums II running Windows NT to 10/100 unmanaged switches everywhere. It was tough to put it to a somewhat "OK" standard but I got a nice experience on the 2 1/2 years I was there.

Improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.