Jump to content



Photo

[Split] Exchange 2013 Talk - Love it or Hate It?


  • Please log in to reply
13 replies to this topic

#1 +Tech Greek

Tech Greek

    Neowinian

  • Joined: 22-October 08
  • Location: Shreveport, LA

Posted 16 February 2014 - 05:54

I migrated 2010 recently but luckily had none of these issues. Now im getting ready to do a 2013 migration again. Lol

Edit-
Actually after reading it a bit I did have that issue on a few users but figured that out pretty fast based on my experience with my lab setup.

Experiences like that is why I'll miss technet! Learned all my Exchange and Server 2012/R2 stuff by doing and breaking and fixing stuff!

 

Good luck with the Migration, it's been a nightmare for me. I thought 2007 to 2010 wasn't smooth boy was I wrong!




#2 Roger H.

Roger H.

    Neowinian Senior

  • Tech Issues Solved: 20
  • Joined: 18-August 01
  • Location: Germany
  • OS: Windows 8.1
  • Phone: Nexus 5

Posted 18 February 2014 - 00:26

Good luck with the Migration, it's been a nightmare for me. I thought 2007 to 2010 wasn't smooth boy was I wrong!


I think you jinxed me :p

One whole weekend of issues. From EAC not connecting (HTTP 500) to SSL issues (have a CA cert I imported) SMTP crashing (451 errors) and that was just the beginning. I finally got all that sorted out then the server started getting RAPED by IIS Worker Processes (also had some noderunner.exe issues but those were expected since they have to index everything for the fast search thing). Last thing was just losing connectivity with my Exchange Online Protection (EOP). Good thing I had that trial too as having to reboot services and the SERVER 10 times each would have probably lost some emails here and there.

I'm not sure if it's a Server 2012 R2 thing - it's not technically supported yet or if it's a Generation 2 Hyper-V issue but slow as hell, even with 8GB of RAM (to start, was gonna add more).

Now I just created a new VM on a different host (not enough RAM to run both on the same machine - plus 2 other VMs for SQL). This one was running Server 2012 so only Generation 1 was created and i'm not moving mailboxes off that crazy one to the new one, gonna take a while to sync. This is just a staging box though as i'll recreate a new VM and set it up properly with the storage and such and proper host for RAM and then move the mailboxes back again. LOL

But it was either that or having issues for weeks trying to figure out what's wrong with the busted one!

#3 OP +Tech Greek

Tech Greek

    Neowinian

  • Joined: 22-October 08
  • Location: Shreveport, LA

Posted 18 February 2014 - 08:54

I think you jinxed me :p

One whole weekend of issues. From EAC not connecting (HTTP 500) to SSL issues (have a CA cert I imported) SMTP crashing (451 errors) and that was just the beginning. I finally got all that sorted out then the server started getting RAPED by IIS Worker Processes (also had some noderunner.exe issues but those were expected since they have to index everything for the fast search thing). Last thing was just losing connectivity with my Exchange Online Protection (EOP). Good thing I had that trial too as having to reboot services and the SERVER 10 times each would have probably lost some emails here and there.

I'm not sure if it's a Server 2012 R2 thing - it's not technically supported yet or if it's a Generation 2 Hyper-V issue but slow as hell, even with 8GB of RAM (to start, was gonna add more).

Now I just created a new VM on a different host (not enough RAM to run both on the same machine - plus 2 other VMs for SQL). This one was running Server 2012 so only Generation 1 was created and i'm not moving mailboxes off that crazy one to the new one, gonna take a while to sync. This is just a staging box though as i'll recreate a new VM and set it up properly with the storage and such and proper host for RAM and then move the mailboxes back again. LOL

But it was either that or having issues for weeks trying to figure out what's wrong with the busted one!

 

Jinxed you? HAH! I jinx my self...

 

So check this out, tonight my DAG (Exchange 2013 array/coexistence with 2010) sitting there chugging along, life's great - oopsie run out of space on one server. No big deal right? I decided to turn Circular Journaling on temporarily so I could expand the HD.

 

WRONGO

 

The thing DIDN'T DISMOUNT (which I'm HIGHLY HIGHLY convinced is because of some sort of SICK JOKE of a user interface they call the ECP now)...but wait it did dismount, sort of. Corrupting my entire DATABASE into the point of no return - along with it my backup from earlier that day (don't even ask me how that one happened). I ended up only having a backup for this database from the PRE-2010-2013 move so that wouldn't work and would have been a waste of time. Hey did I mention the DAG grabbed my corrupted DATABASE AND REPLICATED IT! It thought it was just peechy (but won't run a /p or /d on it due to an invalid duplicate key error on their terrible jet db).

 

I ended up having to recreate a blank EDB and export everything to a PST from the OST inside of Outlook...yeah, that's not so easy when everyone's pushing 5-20 GB Mailbox sizes. 

 

If it makes you feel any better I'm running all of mine on Server 2012 R2 with 32GB of ram per VM (GEN2) (with PLENTY of overhead left) and it still randomly complains about low memory on every server.

 

Moral of the story? Triple redundancy from now on with JUST a random mailbox update every two hours versus waiting until the server backs up every night. The server seems to run stable when the damn thing finally boots but this night has been insane to say the least.

 

Disregard any grammatical/spelling/sanity errors, I just pulled off a 48 hour shift of beast mode.



#4 OP +Tech Greek

Tech Greek

    Neowinian

  • Joined: 22-October 08
  • Location: Shreveport, LA

Posted 18 February 2014 - 18:25

So the ultimate lesson learned here...change the OOB concurrent connection values.

 

They apparently were spiking a DDoS attack at me (Chinese IP's) and it would trigger the DB to go offline due to the low limit out of the box I guess preventing it from killing the server CPU/RAM wise (and these were brief, so brief that it wasn't enough to alert me) - this coupled with the circular logging just happened wayyyy too close at the same time and one too many DB remounts yesterday grenaded it.

 

Luckily I used cached exchange on every user since the backup was corrupt so I was able to recover everyones information right back to exchange and fix the exploit/block the range.



#5 OP +Tech Greek

Tech Greek

    Neowinian

  • Joined: 22-October 08
  • Location: Shreveport, LA

Posted 18 February 2014 - 19:06

Scratch this, split view fooled me :D

 

HATE IT! Eventually they'll fix it but right now it's too young to let people experience advanced issues with no possible solutions.


Edited by Tech Greek, 18 February 2014 - 19:14.


#6 Roger H.

Roger H.

    Neowinian Senior

  • Tech Issues Solved: 20
  • Joined: 18-August 01
  • Location: Germany
  • OS: Windows 8.1
  • Phone: Nexus 5

Posted 18 February 2014 - 19:34

Hehe, figured that would get you but didn't want to cloud up the other topic with our banter. Figured other people might want to chime in too so......

But for your EAC stuff, I actually kinda like it.... I don't miss ECP so much from 2010. Mostly everything seems well laid out and I found all the options pretty easy when I installed 2013 for the first time.

Sounds like you had a crazy 48hrs though. I enabled circular logging on my "temp" server for now since the machine it's running on doesn't have much storage even though I could just move just the storage to the old Hyper-V Host. Then again 20GB mailboxes? I thought mine were huge with 5-7GB!. You have 90users too so even more than me.

Oh and Cached Exchange FTW! :p

And what's the OAB currently set to? Didn't even know that was a thing :p. Guess I should look it up and see what's a good number based on users to lower it to?

So far my new server is working just fine too so taking my time with creating the new one. Doing updates, testing and everything before even installing Exchange even. Once I do that then i'll let that ride for a while too before moving mailboxes over. :p

#7 OP +Tech Greek

Tech Greek

    Neowinian

  • Joined: 22-October 08
  • Location: Shreveport, LA

Posted 20 February 2014 - 09:38

Hehe, figured that would get you but didn't want to cloud up the other topic with our banter. Figured other people might want to chime in too so......

But for your EAC stuff, I actually kinda like it.... I don't miss ECP so much from 2010. Mostly everything seems well laid out and I found all the options pretty easy when I installed 2013 for the first time.

Sounds like you had a crazy 48hrs though. I enabled circular logging on my "temp" server for now since the machine it's running on doesn't have much storage even though I could just move just the storage to the old Hyper-V Host. Then again 20GB mailboxes? I thought mine were huge with 5-7GB!. You have 90users too so even more than me.

Oh and Cached Exchange FTW! :p

And what's the OAB currently set to? Didn't even know that was a thing :p. Guess I should look it up and see what's a good number based on users to lower it to?

So far my new server is working just fine too so taking my time with creating the new one. Doing updates, testing and everything before even installing Exchange even. Once I do that then i'll let that ride for a while too before moving mailboxes over. :p

 

I've been getting more and more used to it now, I still find my self doing things that should be much quicker but the ability to do so from anyone's machine (say, add a user) while on site so I don't have to break out the VPN/Laptop (yes, that 8 second SSD boot time is a killer!  :woot: ) has saved me some time already

 

My connections were set on some ridiculously low setting (20 I believe) and I think that it got confused when it tried to coexist with the Exchange 2010 servers (should have been minimum of 200).

 

I've got all of the 2013 on a DAG group now, I've separated the CAS/Mailbox role even though Microsoft is kind of straying away from this for ease of spreading it across the hosts that are a little more confined on ram.

 

Setting the new DAG on 2013 was somewhat of a nightmare, why you have to pre-stage the DAG in AD is beyond me and the permissions deal that they haven't fixed yet that you have to apply (that should have all been automatic as far as I'm concerned). Now that the DAG is fully setup and I've tweaked the cluster settings for it it's been pretty much bullet proof in terms of keeping everything in sync.

 

Some issues I ran into - one of my CAS/MB Combo servers decided to turn off the page file...so it was screwing with IIS, etc - screwed up the content index on one database pretty bad apparently that I had to rebuild. I'm not sure why the pagefile setting got changed (or how) but I know to double check it now.

 

Now I've got to debate on when I'm going to do that wonderful .local to web domain AD transfer. Hopefully with the reverse proxy I don't need to worry about it anytime soon.

 

So, cliff notes:

 

1) Check the Page File

2) Start building a list of frequently used PowerShell scripts that you'll use

3) Server Settings need to have the Network Adaptor selected (and not use all) or else you'll deal with terrible spikes in lag

4) Don't chance a circular log (no matter how temporary) it may be unless you have JUST backed up - not the night before.

5) DAG had to be pre-staged, file share permissions for witness had to be set, Exchange System had to be added to admin account

6) Use DHCP for the cluster - seems more stable for whatever reason in terms of issuing commands now (and you don't have to worry about assigning statics)

7) Run the verification test on the cluster to be happy

8) Exchange 2013 is ok now that it's acting stable and I've fixed all the little bugs; I dread any disaster scenarios - hopefully all of the DAG/Cluster/Redundancy they've built in will be seamless - I'll be testing it this weekend.

 

Happy Belated Birthday  :fun: , I saw yesterday the confetti but I was dealing with a TON of people calling in from the road that I had to pull OST into PST and then delete profile and reimport it (rather than dealing with recovery on my backup as it was corrupted and wouldn't just work on a restore, would have had to remote in and fix profile anyways due to the new EDB creation). Since Sunday night I've had 3 hours of sleep, I just finished the last MDB server and started a migration so I'm about to crash...crash...crash... :sleep:



#8 jamieakers

jamieakers

    Neowinian

  • Tech Issues Solved: 1
  • Joined: 08-January 10
  • Location: London, England
  • OS: Windows 8.1
  • Phone: Dell Venue Pro

Posted 20 February 2014 - 10:25

I hear you ... have to do software testing against Exchange 2013 pretty much every day.  It feels like a Vista-esque release of Exchange, it simply doesn't have the performance or stability that usually comes as standard with Exchange.

 

Oh and don't get me started on the ECP.  Nicknamed the "Exchange Crash Panel" round these parts.  Generally we PowerShell everything, however if we have to use the ECP we've found it's a whole load more stable on Chrome than IE or any other browser.  Go figure .. !



#9 OP +Tech Greek

Tech Greek

    Neowinian

  • Joined: 22-October 08
  • Location: Shreveport, LA

Posted 22 February 2014 - 08:00

I hear you ... have to do software testing against Exchange 2013 pretty much every day.  It feels like a Vista-esque release of Exchange, it simply doesn't have the performance or stability that usually comes as standard with Exchange.

 

Oh and don't get me started on the ECP.  Nicknamed the "Exchange Crash Panel" round these parts.  Generally we PowerShell everything, however if we have to use the ECP we've found it's a whole load more stable on Chrome than IE or any other browser.  Go figure .. !

 

Exactly except I've had issues with PowerShell randomly losing the ability to communicate randomly giving transit errors [which return people running into the same thing in the partner forums with NO solutions or explanations as to what any of those error messages actually mean] (which seems to be fixed with the DAG on 2013 - ironic because it thought it needed a DAG OOB just to make a DB and threw up an error even though no DAG had ever been made).

 

Chrome is what I've used since day one on it but I did try IE to manage it and I noticed there was issues randomly with clicking things after I had the ECP open a while. Chrome has none, go figure.



#10 +Audien

Audien

    Software Eng.

  • Joined: 30-December 03
  • Location: Seattle, WA
  • OS: Windows 8.1/Mac OSX
  • Phone: iPhone 5S

Posted 24 February 2014 - 02:47

I'm not familiar with HA or ECP.  Though if anyone is having problems with CAS15 I can help with that!  



#11 Roger H.

Roger H.

    Neowinian Senior

  • Tech Issues Solved: 20
  • Joined: 18-August 01
  • Location: Germany
  • OS: Windows 8.1
  • Phone: Nexus 5

Posted 25 February 2014 - 18:59

Welp - Exchange 2013 SP1 has been released!

 

http://www.microsoft...s.aspx?id=41994

 

also the UMLanguage Pack

 

http://www.microsoft...s.aspx?id=41997

 

Get the downloading! Since i only built a test server i'm wondering if i should just build a new one again from scratch with SP1 instead of upgrading.



#12 +Audien

Audien

    Software Eng.

  • Joined: 30-December 03
  • Location: Seattle, WA
  • OS: Windows 8.1/Mac OSX
  • Phone: iPhone 5S

Posted 25 February 2014 - 22:42

Welp - Exchange 2013 SP1 has been released!

 

http://www.microsoft...s.aspx?id=41994

 

also the UMLanguage Pack

 

http://www.microsoft...s.aspx?id=41997

 

Get the downloading! Since i only built a test server i'm wondering if i should just build a new one again from scratch with SP1 instead of upgrading.

 

I strongly encourage everyone to try out SP1.  Everyone has put so much effort in to make it an awesome release.  



#13 Roger H.

Roger H.

    Neowinian Senior

  • Tech Issues Solved: 20
  • Joined: 18-August 01
  • Location: Germany
  • OS: Windows 8.1
  • Phone: Nexus 5

Posted 28 February 2014 - 00:04

Running now and no issues other than some silly ones that haven't been fixed yet (event log stuff, nothing "real" yet).



#14 jgreene

jgreene

    Resident One Post Wonder

  • Joined: 11-July 14

Posted 12 July 2014 - 00:04

Hate it.  Please keep i mind I come from a mail system not so mired in hardware and AD.  Seriously, why anyone would choose this mail solution at the Enterprise level is truley staggering to me.  If you've only ever know Exchange I guess I get it. If not you will know that it requires more hardware, more admins, is more complex, less flexible and then to top it off, it has to run on Windows!!!  Don't get me wrong, i'm in for the long haul but this is truly 2nd if not 3rd rate.  The positive is we can thank MS for job security!





Click here to login or here to register to remove this ad, it's free!