Tag Archives: “performance”

interesting things i see on the internet – 27/01/2014

first of all, you should all be planning your SP3 upgrades, if you haven’t started already. MSExchange.org are starting a new series this week on this very topic, so as well as reading my earlier blog post on this, you should read their article as well.

Exchange design:

Here are some nice test lab guides/posters on cross product solutions with exchange, lync and sharepoint, and here’s a brief (very brief) article on setting up an exchange 2013 lab from Steve Goodman.

Exchange troubleshooting:

I’ve seen a wonderful script for troubleshooting unexpected database growth. This script will snapshot a database and compare it to previous snapshots, and then tell which mailbox is growing, by how much and how many items. Like using exmon, but about a million time easier. I heartily recommend that everyone has a good play with this, so that when you come to use it in anger (and you probably will), you know exactly what you’re doing with it.

This looks like it may save some of you some pain in the near future; the right way to create additional receive connectors in Exchange 2013.

An old post, but an interesting one – do you have sleepy NICs? A common cause of databases moving around unexpectedly in a DAG. We’ve got a couple of customers experiencing this, and we’ve checked that this isn’t the case, but it would be great if people would check again. This is one for the best practices document, i think.

The Romanian exchange support engineers have suddenly become active on their blog, after years of very occasional posting. There’s a couple of pretty detailed posts covering some interesting troubleshooting issues up there at the moment. The mailflow troubleshooting guide is good, if brief.

Rhoderick Milne has a great post on mailbox quarantine that’s got some great hands on advice on configuration, which may explain to you why some users cannot reach their mailboxes.

Exchange general:

Here’s an interesting article on the new restrictions on upgrading the database schema in Exchange 2013. Note that’s the DATABASE SCHEMA, not the DIRECTORY SCHEMA. In exchange 2010 the database schema of each dag member upgraded as soon as the service pack/rollup was applied, making it tricky to then move databases around until all the nodes were on the correct version. In 2013 the schema won’t update until all members of the dag are at the correct software level to support the updated version. Why is this good? Because it means we are less likely to get in a situation where nodes get stuck on the wrong version and are unable to support an active database. It does happen. Twice, in my experience – 2007 was particularly prone to it.

As usual Tony Redmond has a batch of interesting posts;  how exchange 2013 measures and monitors server healthten predictions for exchange in 2014, calcualting client access licenses, service packs and cumulative updates, and the reappearance of powershell command logging. yay. There’s a lot of other posts there, all of which are great. You should read them.

Here’s a great way to start reporting message flow statistics – how many messages are generated, NDRs and so on, using the ExLogAnalyzer.

The Redmond Interoperability Plugfest 2013 has a video on MAPIHttp, the replacement for RPC over HTTP which i mentioned in the last mail. There’s also one on exchange 2013 protocols and one on outlook 2013 protocols. They’re not long, which is probably just as well. Soo… sleeepy…

Msexchange.org are starting a new series on monitoring exchange 2013 with scom 2012. Given that we are sooo up on monitoring exchange 2010 with SCOM 2007 r2, we should probably start reading this now, yes?

They’ve also got a new series on transport high availability in exchange 2013, which looks like it may be useful  – shadow redundancy and safety are in there, somewhere.

Here’s a really nice script for automating exchange mailbox audit logging. Remember to keep an eye on your disk space. What does the mailbox audit log contain? Who accessed a mailbox, what was deleted, if mail was sent using a “send as” permission and lots more. Of course you want to keep this information.

Tony Redmond (again?) has published an article on managing activesync partnerships for multiple devices on his personal blog.

Core general:

Perfmon incorrectly calculates disk latency in windows server 2008. If we don’t apply this hotfix, then we can’t trust what perfmon is telling us; given that disk latency is a major cause of poor user experience, you really need to get this installed.

MS have published a complete and updated list on microsoft product virtualization – what is supported and what is not, here.

there is a new MATS tool released for analysing the storage stack on windows 2012 and 2008. What’s MATS?

It’s not great, yet, but it’s highly promising – the solutions node in technet. How-to guides for all things, eventually. In the meantime, single sign on federation in hybrid environments…

Office 365:

Ali Larter demonstrates how office 365 stops her from trashing banks, cop cars, hotel rooms and so on. Save the cheerleader an that.

MS is developing a series of Test Lab Guides on Hybrid solutions – the Office 365 trial subscription guide is discussed here. Is it great? Well no, but it’s just part of the whole hybrid stack of test lab guides – see also the solutions stuff above – if MS manage to pull this off, it’ll be awesome. If they don’t, well, hopefully it will have given you some ideas. Here is the first step on the stack, the windows 2012 configuration test lab for public cloud technologies.

The EHLO blog has some useful, if basic, guided walkthroughs on mailbox and folder sharing scenarios. It’s Nino Bilic – if he thinks it’s worth writing about, it’s probably worth you reading about it.

it’s windows azure jump start week at the Microsoft virtual academy. live video from 8pm til midnight every night from now until Friday. or you can wait a week or so until the recordings get posted.

MS have published a useful collection of KB articles on troubleshooting common Office 365 issues.

and that’s it. my inbox is now clear. time to move to Asana. not.

Exciting things I have seen on the internet November 18 2013

Performance Tuning Guidelines for Windows Server 2012 R2– ok- it’s a little early for most of us (like, two years too early), but it never hurts to get a jump on these things. This is pretty comprehensive (meaning looooong), and I’d suggest you read it BEFORE you’re up to you’re a**e in alligators.

AV Exclusion Recommendations  here you go – everything. With a cherry on top. There is now NO EXCUSE for a server to be incorrectly set up. This document makes a great Christmas present, by the way. All my family are getting hardcopy.

Dual IPv4/IPv6 name resolution this is very interesting. Really. For those of who think you can ignore it because you’ve got IPv6 disabled – Microsoft recommend  you DO NOT disable IPv6 –  stop being naughty.

Clint Boessen’s 5 minute guide to office 365 migrations. If you’ve only got five minutes, but want to know about office 365 migration… this’ll do it.

There’s a new version of the MBSA out. You know what it stands for, you use it regularly, in fact, I’m shocked you haven’t told me that it was released. I guess it just slipped your mind, yeah?

Last month I told you all about the new on-prem Outlook Connectivity Guided Walkthrough tool – well the Ehlo blog has finally written a post on it. So slow. So very slow.

Here’s a really good post by Andrew S Higginbotham on troubleshooting Transport agents. We’ve got at least one issue like that rumbling along at the moment.

Fixing failed content-indexes from Paul Cunningham

Tony Redmond has once again written a bunch of interesting and useful articles:

In fairness, there’s very rarely a post on his site that isn’t worth reading, for one reason or another.

Exchange 2013 CU2 doesn’t let you have more than 50 databases, despite it claiming otherwise. It’ll be fixed in CU3, apparently. I’ll let you have the list of things that are BROKEN in CU3 when I get it. Sigh.

Talking of CU3, it’s late. I could tell you why, but etc etc. let’s just say something cropped up at the last minute. Again. It’s getting predictable and boring, now.

At what point can I blame the storage?

I quite often find myself wishing a problem would go away. I get desperate to hand off the whole thing to, say, a SAN engineer and be done with it. At what point then is it possible, when troubleshooting poor storage latency,  to do this?

Say i’m looking at some pretty shoddy secs/write on a LUN. I can say to the storage guy “i don’t like your storage”. He then tells me that the secs/transaction he sees are fine. Is there any way to get just a little bit more info out of the box to narrow down the problem (or slope my shoulders just a little bit more?)?

yes there is.


As far as i can make out, if you follow that article you can get a measure of how long it takes the storport driver to get a response from EVERYTHING ELSE. So, if you’re seeing 50ms/read in the perfmon, but it never goes above 30ms to service a request, you might want to make sure you’ve got the latest version of storport on your system. Conversely if (like me) you’re seeing 500ms request times in storport, you get together with Mr Storage and go yell at the network elves.