Outlook 2007: PST and OST files may be up to 20% larger after the February 24, 2009 Cumulative Update

As per: http://support.microsoft.com/?kbid=968413

The February 24, 2009 cumulative update for Outlook 2007 includes changes related to the general reading and writing characteristics of the .pst/.ost file so that less is demanded from your hard drive to perform the same general operations in Outlook. For instance, the file flushing process was optimized so that it interacts more efficiently with operating system and disk-write caches. There is also a change in the way new data is written to the .pst/.ost so that the data is saved much closer together on the physical disk. As a result, there is less physical work required from the drive, thereby speeding up the time it takes to write the same amount of data to the drive. However, a side-effect of this data writing algorithm is that .pst/.ost files will be about 20% larger than before the February 24, 2009 cumulative update.

Most people will not have noticed the difference as many corporate clients use Exchange & Outlook so never see the OST files but those with PST files may have. For what it’s worth with the cost of storage being so low I would put up with the storage increase for the performance gain.

Growing pains

When I was a kid growing up I used to get all sorts of aches and pains, when I asked what they were more often I was told these were ‘growing pains’. With the sharp departure of my teenage years meant I saw the last of these ‘growing pains’ but I didn’t realise that they would again come back to haunt me.

Recently business has been good, I mean very good.  The last couple of weeks has seen my week more than full of work which has been absolutely great as more work means more money of course.  The difficult thing with this type of influx of work is trying to manage this influx of work with the existing work as well as fulfilling the expectations of new clients.  This in itself is not an easy task as predominantly most new clients have a major or a significantly challenging issue which requires urgent attention.  For example, I had a customer call through to the “on-call” support phone at 8pm last Friday night asking for someone to go out straight away as he’d had no email for 2 days because the broadband had been changed over. This meant that despite being on a Exchange installation over the weekend it needed sorting out, 3 hrs later I was back home again. Going out at that time was not an issue, after all I love what I do, the trick is managing all of the expectations that come with this. This type of work also throws the entire work schedule into disarray and some tough decisions need to be made as to how to organise the week. 

Our clients have always understood that if ever there is a high priority job because of an emergency that we pull out all the stops to sort it out. Which essentially means that if the scheduled work is non-critical, ie. routine visit, then we would ask to rearrange the work if an emergency comes up. They agree with this policy as this is the type of service that they would expect so when it is their stuff that needs rearranging, once we explain the situation they will work with us to rearrange.  This policy has been a tremendous help both in the past as well as in the present but sadly there is not always enough hours in the day to get things done. As it stands we are getting close to the point where we will have to consider bringing another body on board to cover the new work but we need to ensure that the work coming in is of a consistent level before making the leap.  So how do I decide what’s not so important? Here’s some simple guidelines:

• Is this job a fault/failure?
  Some site visits may not be as a result of a fault or failure they may be in-fact a scheduled site visit.
• How many people does this situation affect?
  If the issue that you are responding is affecting more than one person then it will be difficult to “bump” it but if it is only affecting one user then it may be possible.
• Can the client wait until tomorrow/the next day etc?
  If it looks to you like the job can wait, them ask your client. We always explain to the client why we need to change the appointment and explain to them that in case of emergency we always endeavour to pull out the stops and would do the same for them. This usually sways things.
• How important is it really?
  This is a tough one as you may not see a signature in Outlook as being critical but to the user you can be sure that it is.
• Can it really be done remotely? 
  This should be a no-brainer as it should have already been tried but sometimes the client will ask for a site visit and you just agree.

The most important thing is to engage your client and explain why, at then end of the day no matter how much preventative maintenance we perform, no matter how much proactive management we do there will always be something that breaks/goes wrong. We always make a point of explaining to clients that we have the “bump” policy that says that if an emergency arises then we will attempt to reschedule any non-critical site visits so that we can get the maximum man-power on to an emergency situation.

Hope you all have a great weekend and better week.

- Rob

The best laid plans

Things have been a little quiet on the blog recently.  This was mainly down to a project that was being planned for a client.  The project consisted of a transition from a Windows 2000 Active Directory running Exchange 2000 over to a new Exchange 2007 server running on Windows 2008 server and to make things a little more fun there was a 110 mailboxes to move.  Not really a major job as the process is more about step-by-step processing but none the less it still needs to be fully planned to ensure that all the bases are covered.  The reason for the work being done in the first place is mainly due to the current server being 8 years old, the HP LC2000 had really served them well, but more importantly one of the drives in the RAID5 had failed.  So while the hardware was ordered from Dell and awaiting deliver we put together a full plan to handle the transition from the old server on to the new one. As you may know, the transition can not be done directly as there needs to be a Windows 2003 domain controller in the active directory so the plan would be to make the new 2008 server a directory controller on the domain to overcome this issue.

The first job when we got onsite was to get a full backup of Exchange as the client had not done this for sometime due to technical issues they had faced with the tape drive but had not dealt with previously. While the backup job was running and copying the data to a USB drive I set about configuring the new server for use. Unfortunately Dell had seen fit to pre-install the server with the 32bit version of Server 2008 despite knowing it would be installed with Exchange 2007 so I had to get the server re-installed with the correct version.  Then we had to prepare the directory for the insertion of the new 2008 directory controller as the forest needed to be prepped in order to accept the server. Once the server was finally running as an active directory controller for the 2000 Active directory it was time to get Exchange 2007 server installed on the new server and integrated into the existing Exchange Organisation. Sadly during the backup of the old server another drive in the RAID5 failed and that was the end of the Exchange 2000 server. In most businesses this would have been a complete disaster however due to the way that they have worked for some time, most mail is saved to PST files on the server, this was a real pain and caused a little hassle but it was not a massive disaster. It was a case of removing the legacy mailboxes from the new Exchange 2007, which were non-existent, and replacing them with new Exchange 2007 mailboxes with the same details using the PowerShell.

Now while the failed drive was always a consideration with this process but unfortunately they did not want any action done on the issue. The fact is that regardless of how much planning is put in to a job, no matter how detailed the project checklist is and more important how much effort you put in; sometimes shit happens! The important thing is not what happens but how you cope with the event and ultimately how you recover from it.

The outcome of this is that the client is now looking to put in our latest BDR solution that gives near-seamless fail over in the event of a server failure for up to 5 servers.

The lesson of the day is do NOT neglect your backups, if they fail or if you fail to do them you are screwed. No two ways about it; ask yourself how long could you business survive without its computers or its client information? Does £100 for an external USB drive to backup your data real sound so expensive now? Contact JPT Solutions to discuss backup and business continuity solutions in more details.

- Rob

There’s a price on yer head

Some of you may have read that latest news that Microsoft has put out a $250,000 reward to find who is behind the Conficker worm that is said to have infected as many as 12 million computers.  They are doing this because it has taken the firm view that the creation of the Conficker worm as a criminal act.

This may seem like a bold move but it is not the first time they have done this. In 2005 they paid $250,000 to two individuals who helped identify the creator of the Sasser worm.  Rewards were also offered of $250,000 for the creators of the other three major computer worms Blaster, MyDoom and Sobig however the authors of these were never caught.

In reality this amount of money is a small drop in the ocean for a company like Microsoft but in doing so they are trying to send out a strong message to the authors of such worms.  They are simply saying that they will not sit idly by while the creators wreak havoc on their clients systems.  In reality the fact of the matter is that regardless of what Microsoft or Microsoft Trustworth Computing Group offer it seems that it will offer very little in the way of a deterrent for such authors as it is a challenge that they enjoy rising to.  What it does do is help their clients to feel that they are doing all they can to try an prevent such hassles returning in the future.

The worm itself infects a computer that is not fully up to date with the latest updates from the Microsoft Update website. If you are in any doubt then the best action is to visit the Microsoft Update website and apply all the latest critical updates.  Once this is complete continue to revisit the site until you are told there are no further critical updates. You should also ensure that your Anti-Virus software is fully up-to-date, if you don’t already have one then visit either AVG or aVast who both provide free versions for home use.

Network woes

Today I was called to a clients that had added some new RAM to a Windows XP pc but could not access the network any longer.  What was happening was it showed a connection was in place but it would not obtain an IP address from the network nor would it workwith a fixed IP address.  The first thing to check was the physical side of the networking such as cables, network ports etc, but still to no avail.  Tried a known working connection from another PC and drew the same blank result which started to look like cooked OS which was not something I wanted as the look in his eyes said he wanted it sorting today.  So I tried a new network card in case the internal adapter was toasted but stil no good. Next option was going to be trying repairing the TCP/IP stack using the time old command of:

netsh int ip reset c:\ipreset.txt

Still no joy, now things were really starting to look bleak as this was an older unit that shipped with an OS CD Rom and not the good ol’ recovery disks.  Then with a burst of inspiration I decided to try the winsock stack, as it was possible this was damaged as a standard “ping” command produced some very odd results.  The command, as before, was:

netsh winsock reset

This was followed by a restart and cries of joy (inside of course, outside it was “there we go all sorted”).  This is not a command that I have had to use before, many a time the ip reset has been used but no this one.

Anyhow, you learn something new everyday!

Fixin’ stuff

Over the weekend it was time to re-install the work laptop, which isn’t too much of an issue however the data need to be taken off first.  So the first job was to try and find somewhere to dump about 40gb of data, nothing too major but the state of modern computing shows that your average desktop/notebook typically has more storage than servers.  The best place to put this data was going to be on the host for our three virtual servers but for some reason, following the installation of VMWare 2.0, network access to this server stopped altogether.  The server itself could get out but nothing could get back into the server itself.  What transpired after a number of hours of testing was that during the upgrade to VMWare Server 2.0 it changed many of the policies in the Routing and Remote Access Service which meant that there was not NetBIOS access to the host server but the virtual servers were still accessible which was more confusing.  The end result was to downgrade VMWare Server to a previous version and then to disable the Routing and Remote Access Service.  Following a reboot of the host everything came back on fine and network access to the host was fully restored.

All that was left was to reinstall Vista back on the laptop which went very smooth, so much so that I now have a dual boot to Windows 7 Beta configured so that I can start and give it some serious testing.  More to follow on this one.

Where does all the time go?

As I posted recently we are in the process of implementing Autotask which is not a short process however even in this short period of time it is proving to be well worthwhile.  So what have I learned from this process so far?

Business automation is a not a simple process. This quite simply is because the processes which we have in our businesses, regardless of size, are complex and each step is usually dependant on a series of processes which have been completed beforehand. The next thing which has become apparent is that there is a great deal of work which we undertake which is “rationalised”.  You may wonder what the hell I am talking about so lets give an example;  A client calls to say that they have a problem with a notebook and they need us to take a look, we arrange to visit the client at site and take a look.  This visit, only including time onsite, takes 1.5hrs as diagnostic work is performed onsite but the unfortunately the unit has to be removed from site and reinstalled.  It is then brought back to the office to be reinstalled, the client did not have the recovery disk, this process takes a further 3 hrs including updates, drivers etc.  Then the unit is returned to site for final configuration and demonstration to the client to ensure they can see it is working which takes a further 1.25hrs.  All in the process takes a total of 5.75hrs however we know that the client is not going to pay that much for the work to be performed so we only charge for 3hrs.  Some may see this is as being too generous by not charging for this time but we believe in being fair with our clients, sadly the client still thought this was expensive but then you can’t please everyone.

The moral of this is that once you start to account for all of your time then certain things become clear and lessons to be learned.  You may not like these lessons they will ultimately benefit your business, while you still may continue to make these time concessions they will be done knowing what time is being given for free and what is being charged for.

Making work-flow

Before we, as a business, moved into managed services I looked into how it was being done and the various software that is typically used to implement a managed services practice.  There are two core elements that were being repeated over and over again and these were:

• Desktop/Server agents – these will typically be used to provide the necessary features to make delivering managed services easier and more importantly efficient.
• PSA Software (Professional Services Automation) – this is the software that will be used to manage your deliverables and help you to ensure that you are delivering on your promises to clients.

The first point that we looked at was the management software that was on offer.  The key player at this time was Kaseya who at this point had a very large share of the desktop/server market.  The software itself is very compete as a package but we found that it did not lend itself well to small deployments in a cost effective manner.  This led our search deeper and it was then when we found Zenith.  At this time a relatively new player but none the less a very powerful toolset and extremely cost effective.  The product also did not have a steep learning curve which meant that the tech’s were able to quickly get on board with the product.  This meant that we were sold on the product that we’d use to manage our clients site’s, next it was on to the PSA.

The next hurdle was to be the decision on what PSA software we would be using to manage the clients.  As with many MSP’s we initially decided to start by running a relatively manual system by keeping notes which involved a combination of both spreadsheets and completed paperwork that is then filed.  Within a short period of time we found that we started to outgrow this manual system so we then had to undergo the process of looking a tools that would be able to manage this a little cleaner.  The first product that we tried was osTicket, this worked great for simple ticketing based on email transaction as well as web based entry.  Now to be fair this worked fantastic at what it did but we needed more from it, we required more integration of the fringe elements such time tracking for onsite work as well as site documentation.  Next we looked at Shockey Monkey, this is a project being developed for the community Vlad Mazek (OwnWebNow) which is shaping up to be an amazing project but unfortunately we were quickly being over taken by our needs which meant that we needed to look for a ready made package.  After much searching we have finally settled on AutoTask.  This will truly help our business moving forward as the whole workflow will be managed from end-to-end.  From the moment the call comes in a ticket is created.  From here and work done is added to the ticket, if onsite work is required then a site visit can be added to the ticket.  For any work done on the ticket it is allocated a code which allows us to govern what charges are applied and if they have a Managed Service Agreement then the time is allocated to the account and is tracked through the account.  This means that we can see very quickly that our accounts are profitable or not.  When the work is completed we can then make any adjustments to charges and the post it for invoicing.  The final step is then to import the invoice into QuickBooks directly.  If terms of the techs onsite, they can now fill in the work as soon as they leave site from a PDA/handheld device which means we have the most accurate information possible as it is still fresh in their mind.

All of this works to make our business run smoother and more importantly make my life easier which I am all for.  Many of the lesson I have learned others have learned and documented before me, so why listen to me?  No idea, but for what it’s worth I didn’t listen but sometimes you have to work things out for yourself

Spam prevention

Recently we had some problems with our DNS provider for the business which caused us to look at moving the domain over to another more reliable provider.  In the end we decided to settle with hosting the DNS with pipex and while they may not have always had the best reputation they are one of the larger players, considering the problems recently meant that we had no email or website traffic we were not able to get caught with that again.  All went well with the transfer as I made sure that all of the DNS records were live before the actual transfer took place so the transition was completely seamless.  However, since then it would seem that the spammer capitalised on one oversight that I made during the transfer; this was SPF (Sender Policy Framework).  For those that don’t know what this is basically it is a TXT record within DNS which defines the hostnames and/or IP addresses which can receive and send mail for the domain name it applies to.  This TXT record is read by any mail server that can support it and should the email not have originated from the correct address then it is treated as spam and discarded as such.  If, however, the mail originates from a valid address then it is processed as normal and passed on to the recipient.  On the other hand if the recipient mail server does not support SPF then this information is simply ignored.  Since I hadn’t got this information handy I was tasked with re-creating the SPF data which for those that have looked it is not the most straightforward thing to do however Microsoft have very kindly created a wizard on their website which takes out most of the work.  The address is as follows: http://www.microsoft.com/mscorp/safety/content/technologies/senderid/wizard/

Using this wizard you are able to enter your domain name and it will process the DNS record to see what settings exist with regard to mail servers and then guide you through the process of creating the SPF data.  Over the past few years I have using a number of tools but this has to be the easiest to follow.  So check it out, the more people that use SPF the less spam will be possible.

As always thanks for reading and if you have any questions then drop me a line.

Linksys woes

Over the weekend I was onsite finishing off and installation for a client of a VoIP PBX system that we had provided.  What they had got was one of IP PBX’s with 4 Linksys SPA941 IP Phones connect to 2 ISDN2e lines so that they had 4 incoming calls.  Due to the nature of their business they get a large volume of calls and the 2 analogue lines they had previously was just not cutting it.  Anyway, I connected the last phone to the system and decided to set the daylight savings time rule on the SPA handsets so that BST would be correct on the handsets.

So I dutifully changed the config on server so that the new correct dates were inserted and then did a factory reset on the handset to force it to get the correct configuration downloaded to the handset……boy was that a mistake!!  What actually happened was that it downloaded the configuration for the server, restarted and when it came up again the display was not working it simply displayed the text logo for the handset.  Now to all intents and purposes it looked as though the firmware had fried which was not something that pleases anyone.  However, after downloading the firmware from the Linksys website it continued to do the same thing and as the display was screwed I couldn’t use the setup option to perform another factory reset.  Anyway, after much backtracking I finally got the old daylight saving time rule from another handset and changed the configuration file on the server along with on the handset and rebooted and finally all was back to normal.  At this time I still don’t know why the rule upset the phone so much but what I do know is that it caused be about 2 hours of hassle the I could have done without on a Saturday.  Once I have tested this out in the lab I will let you all know the fix.