Exchange 2010 Monitoring
Hi,
I have a customer who have Exchange 2010. They have asked me to look into monitoring for them. They already monitor standard things like CPU, Memory etc and some Exchange metrics such as Database sizes, queues etc.
The customer wants to be able to look at Exchange and be confident that it would be fit for purpose if sudden growth occurred (such as an acquisition) and how that could potentially impact the Exchange environment (increased connections, latency etc). Ive
found the counters you can use from technet but this list is extensive. I guess I am looking for real world examples of how I can achieve this and be able to use the data to estimate future growth.
Any help greatly appreciated as always
Load on a exchange system depends on various factors like:
1. User Profile size (message sent/received per day)
2. Number of DAG copies
3. Backup Requirements
4. Dumpster size
5. Storage Infrastructure
6. Hardware / Virtual Infrastructure
7. Number of CAS devices (OWA/RPC/Active Sync)
8. Line of Business Integrations
Every infrastructure is different in many ways depending on parameters described above. Exchange Role sizing calculator should be the benchmark against your deployment. Even Microsoft support would recommend to match the deployment with the calculator result.
I would suggest to use performance monitors for CPU, Memory, Hard disk and Network. Monitor RPC requests, User profiling, Exchange Back Pressure, and Storage I/O load for the current infrastructure. You should always consider 20% of expansion and 10% of
miscellaneous load on the exchange infrastructure.
Always keep CPU / Memory to max 70% usage and ensure DAG database are redistributed equally among nodes. On the CAS side, use Exchange User Monitor to view Exchange CAS load.
Most importantly, keep an eye on CPU/Memory usage on CAS. Disk I/O, CPU / Memory usage on HUB Transport and Mailbox. If the trend is on higher side for many days after there have been change in the head count, you would need to scale up server sizing (CPU
/ Memory) or scale out with more servers.
Let me know if you need more details.
- Sarvesh Goel - Enterprise Messaging Administrator
Similar Messages
-
Exchange 2010 and 2013 Monitoring with SCOM 2012
HI all,
I have exchange 2013 and 2010 in my environment , Iwant monitor both in SCOM 2012,
Can you all please tell can i implement 2013 monitorinng first later can i go to 2010 Monitoring, Please advise in that supported or do i need to go with exchange 2010 monitoring firstThere are no dependencies on the 2010 MP so you should be good to go, see the following:
-
Issue with SLA reports of Exchange 2010 MP in SCOM 2007 R2 !!
Hi,
I have issues with Exchange 2010 MP SLA reporting. Whenever report is run it fails on some or the other dataset(Outage**,SLAbyDAte etc.) When I digged in, I found that it is connecting to DW Exchange 2010 Availability tables + Planned maintainance +
outage table. The tables were not available in the DW, not sure how it got removed. Initially I tried by uninstalling and installing MP(Exchange 2010 & Exchange 2010 reports both) if relevant tables get created automatically but it did
not help and I was still facing the same issue. Then I tried to create the tables manually and check if it could resolve the issue. There was unique ID associated(different in different computer) with the name of most of the tables. I tool it from Exchange
report data source which exists in reporting server. But I could see that table remain empty with no records there and SLA report start giving different issues stating table instances do not exist.
Could anyone help me to restore the SLA report of Exchange 2010 MP ? How should I ensure that data start populating the tables ? I am also helpless here because we do not have correct DW dump to restore back to earlier when it was working fine.
//BRHi,
Based on my research, to correctly install the Correlation Engine, consider the following when you install the Exchange 2010 Monitoring Management Pack:
Determine which server will host the Correlation Engine. While not strictly required, it's strongly recommended that the Correlation Engine be installed on the Root Management Server computer.
Make sure that you download and install the Exchange 2010 Management Pack that matches the operating system for the selected server. You can install the 32-bit or 64-bit version.
Make sure that you run the Microsoft Installer (.msi) package on the server where you want the Exchange 2010 Management Pack and the Correlation Engine to be installed.
After the installation is complete, the Correlation Engine service will have been installed on the server, and the Management Pack files will have been copied to the
System Center Management Packs folder. By default, the installation directory is C:\Program Files\System Center Management Pack.
Quote from:
http://technet.microsoft.com/en-us/library/ee758051(v=exchg.140).aspx
In addition, here is a similar thread for your reference:
http://social.technet.microsoft.com/Forums/systemcenter/en-US/e75e84d9-0c9e-4d83-b3da-45a143757f85/exchange-2010-monitoring-with-scom-2012-correlation-service-issue
Regads,
Yan Li
Regards, Yan Li -
Monitoring Exchange 2010 from external network
I would like to monitoring the following services pf Exchange 2010 from external network / internet
1) SMTP (To confirm the mail has been accepted for delivery)
2) ActiveSynch (Mobile device can login and synch different folders)
3) WebApp (Users can log and access mail)
I have created a script using test-Mailflow, test-activesynchconnectivity and Test-WebServicesConnectivity and running it on server from LAN. I want to monitor the above 3 areas from Internet(external network) to make sure these services are available
from Internet.
We have Barracuda as SMTP gateway, TMG for WebApp and MobileIron for ActiveSynch.
Will i be able to monitor these services from external network(internet) using test commands. What are the alternate ways to monitor the above services from external network.
ThanksWe are trying to build exactly similar to ExRCA. ExRCA is good but it is manual. We would like to build something similar to ExRCA which can monitor exchange services periodically and send alerts.
-
SCOM 2007 to monitor Exchange 2010 Backups
Hi
Is there any way of getting SCOM 2007 to monitor Exchange 2010 backups. I'm using Commvault as the backup solution, was wondering if there's anything i can check for to ensure successful or failed bakcupsHi
DPM will write its own events. You can create your own management pack for SCOM to monitor the events on the Exchange server but SCOM has an exchange management pack you can install. -
How to monitor particular individual Exchange 2010 mailbox
Hi
Is it possible to monitor a particular individual Exchange 2010 mailbox? If so please send me the script or steps to to accomplish this
Donald D'souza (http://donald-scom.blogspot.com/)The Exchange 2010 MP has had some issues of late and is once again unavailable. Hopefully this should be rectified soon. Keep an eye on the Management Pack forum and someone should post there once it is back.
I appreciate the Exchange team are looking for ways to get extra monitoring from SCOM and I'm certainly not saying you can't do mailbox monitoring with SCOM. I'm just saying be careful. I have seen one enterprise write a discovery to discover the
mailboxes they wanted to monitor. The discovery captured data that changed every iteration that the discovery ran. They then wondered why SCOM stopped working, ignoring the fact that they had discovered thousands of objects whose properties changed every
time discovery ran, causing huge numbers of state change events and massive database growth.
If you are looking for in-depth exchange mailbox monitoring then SCOM is not the tool for it. Quest and Sirana do dedicated enterprise tools. If you take a look at the links and say that is what your team want then I'd suggest evaluating those
products. Developing SCOM to deliver that would not be cost efficient.
Hope you don't mind but I'll ask again - what specifically do you want to monitor? Mailbox Size? Number of emails sent \ received? size of message sent? destination? number of attachments? size of attachments? blocked messages? If you need deep detail
along the lines of the following then you really need to evaluate the following:
http://www.sirana.com/exchange/reporting
If you end up writing scripts to do all the above then you will likely severely impact performance of either SCOM or your Exchange server or possibly both.
Sadly (for budgeting) SCOM doesn't necessarily replace specialist tools - it just complements them. Likewise, if you want deep dive SQL then you'll still need Quest or Idera based tools. Probably fewer licenses and better targetted at servers with known
issues.
Cheers
Graham
Regards Graham New System Center 2012 Blog! -
http://www.systemcentersolutions.co.uk
View OpsMgr tips and tricks at
http://systemcentersolutions.wordpress.com/ -
Is it possible to monitor a particular Exchange 2010 mailbox via SCOM if 30 or more emails trigger in 30 minutes in the inbox and alert should be generated for that.
If this is not possible in SCOM. Is there another way we can do it via power shell or any thing.
Gautam.75801Hi,
We may need to create script based rule using PowerShell or VBScript.
Please go through the below links for command get-mailboxstatistics:
Exchange 2010 : http://community.spiceworks.com/topic/122241-scripting-to-count-all-of-the-messages-in-a-mailbox
Regards, Yan Li -
Hi,
We have an issue with our OWA page. We are currently publishing OWA via UAG.
We recently upgraded to Exchange 2010 SP3 and then SP3 Rollup7. Since the upgrade, we are keep getting the following error after entering our credentials on the login page. I've tried with every possible browser.
You do not have permissions to view this folder or page
Strangely enough, the mobile phones are sending and receiving emails just fine, the phones use the same OWA link, so it's not an authentication issue, the phones login into the UAG servers with no issues. I can see this on the Active Sessions screen
on Web Monitor.
I've attempted to connect to the OWA by bypassing the UAG server, so putting in the local OWA address of one of my Exchange servers, it works... so the OWA page is up and running.
No error logs get generated on Web Monitor when we receive the permission error, I think this is because it's past authentication, it's on the Exchange layer.
Any insight would be helpful? I'm assuming something changed on the Exchange side after the upgrade.
Just in case, I've upgraded the UAG and TMG servers to the latest SP and Rollup packets.
UAG > SP4
TMG > SP2 Rollup 5I've found a solution; UAG requires Basic Authentication over OWA. For some reason Integrated Windows Authentication got turned on after the SP3 upgrade.
http://technet.microsoft.com/en-us/library/ee921443.aspx
Turning Integrated Windows Authentication off via the Client Access OWA settings resolved the issue. Though beware, you
have do this on all your Client Access servers. -
Relay issue from Unix across Exchange 2003 OWA server in Exchange 2010 environment
Hi,
I'm trying to resolve an issue. We have one Exchange 2003 server left in our environment. The rest is now Exchange 2010. We are working to decommission this server. Monitoring the SMTP logs, I am working with the various groups to get the traffic
off this server and onto the Exchange 2010 environment.
The issue we are finding is with a work flow. This Java app for eBis sends email to users. This is done correctly through Exchange 2010 and gets delivered from our HUB/CAS servers to the target user mailbox. This email has several links the user must click
for approving or rejecting requests. Upon clicking Approve, a new email window opens (we use Outlook 2010). The TO address is in the format of [email protected] Upon clicking Send, our Exchange 2010 HUB/CAS servers accept the
email, because one of our send connectors has, as address space, *.domain.corp. However, the email address being used it not an alias on any Exchange 2010 mailbox, so it appears Exchange 2010 is sending this email on to the Exchange 2003 server, which
also is an SMTP server. I *think* this server is looking at the address after @ to determine where to send it (ebisserver.domain.corp, which is valid in our DNS), and sends it on to that server, where the java "listener" program intercepts
the mail, processes it and then saves it to a file somewhere.
We are at a loss as to how to get Exchange 2010 to do this instead of Exchange 2003... once this traffic is eliminted from Exchange 2003, I can proceed with decommission of this server. Any help troubleshooting this issue is appreciated.The only reason the Exchange 2010 server would route outbound mail through an Exchange 2003 server is that you have an SMTP Connector defined on the Exchange 2003 server that has a more specific domain than you have on the Exchange 2010 server. You
should be able to see all your Send Connectors (an SMTP Connector on Exchange 2003 looks like a Send Connector in Exchange 2010) by running Get-SendConnector.
Ed Crowley MVP "There are seldom good technological solutions to behavioral problems." -
Exchange 2010 - how to troubleshoot intermittent Outlook slowness
I'm having users report intermittent Outlook slowness 2007. I can't repeat the issue exactly but I have seen it -- you get a message that says "Outlook is trying to retrieve data from <Exchange Server>" and it takes it a noticeable
amount of time. One of the difficulties in troubleshooting this is that I haven't seen it on my own machine, nor can I find the way to replicate the issue consistantly.
We actually just upgraded from Exchange 2003 to Exchange 2010 and never had these issues in the past. I've run a process monitor on the server (which is about 8 times better than our old Exchange 2003 server - 2 quad-core procs and 12 GB of
RAM for about 250 mailboxes) and find nothing abnormal in terms of CPU time or network bandwidth usage (4 teamed 1 GB/s connections). I have also monitored several of the Exchange RPC items, and Client: RPCs but Client:RPCs does not return anything
at all (just 0 the whole time). Are there other things I should be looking at with PerfMon?
I'm at a point where I'm not sure where to take my troubleshooting. Does anyone have any ideas? The fact that I (and most other users I've asked) never experience the issue leads me to believe that it might have something to do with a
local computer issue rather than a server issue, but I'm hearing reports from people with both Outlook 2007 and Outlook 2010. The other somewhat common thread is that
Outlook is NOT running in cached exchange mode, but it wasn't when we were on the 2003 server either and we did not have this issue. Our Exchange 2010 Server is now running the new UM role, however, and we were using Cisco Unity before.
I have already disabled all Outlook add-ins and disabled anti-virus (client, but not Nod32 for Exchange).
Thanks, any ideas would be appreciated.Forgot you're using 2010, the item counts are much more laxed, 100,000 per folder compared to 5000 in Exchange 2007. For your Outlook 2007 clients, can you confirm you're up to date on the last post hotfixes as of Feb 2011? Also I would disable caching
of shared calendars in outlook. Also note that outlook 2010 caches all shared mailboxes by default.
Turn off calendar caching. If you have Exchange in use then Outlook 2007 may well have the caching of people’s calendars turned on. This means that when you open a calendar you have to wait for the cache to populate, which can slow
things down a bit sometimes. To do this:
Go to the Tools Menu and select Account Settings
Select your Exchange Account and press the change button
Press the more settings button
Select the 2nd tab (Advanced)
Un-tick the “Download shared folders” option. Note that this will stop cached Sharepoint folders from downloading into Outlook as well.
Press OK until all options are saved.
Description of the Office Outlook 2007 hotfix package (Outlook-x-none.msp): February 22, 2011
http://support.microsoft.com/kb/2475891
Description of the Outlook 2010 hotfix package (outlook-x-none.msp): February 22, 2011
http://support.microsoft.com/kb/2475877
James Chong MCITP | EA | EMA; MCSE | M+, S+ Security+, Project+, ITIL msexchangetips.blogspot.com -
Office 2010 / Exchange 2010 - Attachments in EML-Emails are invisible
We are investigating a problem that seems to pop up with Exchange / Outlook 2010.
This is the absolutly the same problem that Clemens Wunder wrote about in his tread:
http://social.technet.microsoft.com/Forums/en-US/outlook/thread/afce3bd0-f8a4-419b-986a-0560ebff5f8f
Problem:
We got several eml-files in a Management System and those have attachmentes (word, pdf whatever) included.
If i open those EML Files in Outlook 2010 (this works if the Registry is apropriate) the Email is opening but the Attachments are "invisible"...
If i now save the Mesage from within Outlook 2010 as a MSG File and open this one the attachments are back...
I searched for bout one week now for a solution but i still don't get it why this should be happening.
My Specs:
- Windows 7
- Outlook 2010
- Exchange 2010 (migrated from 2003)
Fact:
The same problem seems also to happen with Outlook 2007 and 2003 since the migration of Exchange 2003 to 2010.
Those EML Files are from within a data management system and are stored as eml since bout 4 years.
Please don't send answers like "Use Windows Mail" - this is not a solution for this matter because I can't.
Exchange / Outlook is used at every customer of our company so WM can't be the solution...
Thanks for your advices and tips
RomanHi Roman,
To make sure I understand things correctly…..
You have EML files generated by another program. Emails have one or more attachments.
You have copied the EML files into your Outlook 2010 profile’s mail folder.
When you open the item in the folder, you have to open or save the item. –You have HKEY_CLASSES_ROOT\Outlook.File.eml.14\shell\Open\command set
to use Outlook 2010 to open the file.
When you open the EML file, the message opens but the attachments are not visible.
In actuality, when you open the EML file, Outlook is saving the file to your “..\..\Temporary Internet Files\Content.Outlook\” location.
To see what Outlook is doing I started a Process Monitor trace,
then dragged a Windows Live Mail EML message with a Word doc attached over to my Outlook 2010 Inbox.
When I open the eml item, I am prompted to Open or Save the item. I opt for Open and the message with attachment is opened with Outlook 2010.
The trace shows accessing the registry to find out what to use to open the EML file and then it shows the attached file being created in
Temporary Internet Files. The attached file is seen,
but nothing happens to it.
28:14.1
OUTLOOK.EXE QueryOpen
jahawk C:\Users\jahawk\Documents\Help me.docx
FAST IO DISALLOWED
28:14.1
OUTLOOK.EXE CreateFile
jahawk C:\Users\jahawk\Documents\Help me.docx
NAME NOT FOUND
28:14.1
OUTLOOK.EXE QueryDirectory jahawk
C:\Users\jahawk\Documents\Help me.docx NO SUCH FILE
From the opened EML file, you click File | Save As | And then save the item as an .MSG file.
When you drag the MSG into your Outlook profile’s mail folder, the mail with attachments is visible.
Looking at the mail folder, the EML item does not have a mail icon associated to it. The item shows no Attachment icon, because it cannot be
read in yet. –You’d need to open the EML file in Outlook to view the message with attachment(s).
Looking at the mail folder, the MSG item is an Outlook mail item and the attachments are showing up.
So, let me ask the following:
How was the item created?
Does this happen with any file that is attached?
In Windows Live Mail, attachments have a lot LESS properties associated to them
Outlook 2010 actually has a PR_ATTACHMENT_HIDDEN property that can toggled to hide an attachment.
What happens from Outlook Web Access? –You will have to right-click “Save Target As” on the attached eml message, first. When you
open the saved “E-mail Message” from the save location, what results?
Is your anti-virus possibly scanning the file? –“..\...\Temporary Internet Files\Outlook.content….”
Anyways, to be honest, I’d have to have a copy of the EML file to be able to (hopefully) explain what is happening.
If I had the file, I would use MFCMapi or OutlookSpy to dump the EML message’s properties after I’ve added it to my mailbox. I’d
then open it and save it as an MSG file type. From there, I’d save the MSG file’s properties off and then compare the 2 sets of properties. And, the item has an Attachment Table that will contain the attached files. Each attachment will have its
own set of properties.
From an Outlook created message with attachment, you could compare the attachment’s properties to the props on the affected message.
From a Windows Live Mail message with attachment, you could compare the attachment’s properties to the props on the affected message.
For the kind of troubleshooting we'd need to do, I would suggest calling and opening a support incident with us. That would allow us to work with you
at a much deeper level. For a case like this, a debug trace of opening the message from in Outlook 2010 would likely be necessary.
Sorry I didn’t have better information for you.
Jahawk MSFT -
Hello. First of all, thank you very much for your attention.
After that, I will expose the problem:
I have a user domain called "comercial" with a corresponding mailbox "[email protected]"
. This user
has assigned multiple emails such as "[email protected]", "[email protected]", "[email protected]" , having as default email "[email protected] ".
On
the other hand,
another domain user
called
"menganito" with it's mailbox "[email protected]" account, also has a few mail accounts "[email protected]", "[email protected]", " [email protected]"
...having
as default email "[email protected]
The user "Comercial" is forwarding
all of it's mails to the user "menganito".
The problem is that "menganito" is not receiving the exact
"comercial"
sender email, but the email which appears by default ("[email protected]").
"Menganito" would like to know
from which email exactly the mails he is receiving have been sent ("[email protected]",
or "[email protected]"
or "[email protected]").
Also, he would like to knowthe email of the original sender (from the outside of the domain) whom the user "Comercial"
have received the email from.
In another type of email servers (Kerio,...)
this is not a problem, because forwarding by
defaultis always showing
the original sender and original receiver.
Maybe the problem comes from the recipient type I'm using.
Should I use another type of recipient? should I use a transport rule?
Any idea ? Thanks again in advance.Hi Simon. First of all thanks for your answer.
This happens with all accounts. I tried with news accounts and different browsers for OWA and Outlook anywhere and happens the same. Is not a cache problem
The problem is
when "[email protected]"(default
and set as default answer
account by
exchange)
receives
a message in "[email protected]" when
forwards to "Menganito"
this message will appear from"[email protected]"
instead of "[email protected]".
This happens with all accounts.
I can understand that i can't avoid
default reply address on the same
domain, but also with different domains happens
the same?
The platform is only one clean and updated server 2012 as dc and another clean and updated server 2012 with exchange 2010 sp3 directly and after rollup 4 for exchange SP3, no previous versions and servicepacks of exchange 2010 was installed before. Before
install exchange 2010 sp3, i installed all the filterpacks (3) and .net 3.5 and all updates that i finded for 2012 server (i tried for 2 days): No more updates for sure. After, in powershell, i put this:
Add-WindowsFeature NET-Framework-Core,RSAT-ADDS,Web-Server,Web-Basic-Auth,Web-Windows-Auth,Web-Metabase,Web-Net-Ext,Web-Lgcy-Mgmt-Console,WAS-Process-Model,Web-Mgmt-Tools,Web-ISAPI-Ext,Web-Digest-Auth,Web-Dyn-Compression,NET-HTTP-Activation,Web-Asp-Net,Web-Client-Auth,Web-Dir-Browsing,Web-Http-Errors,Web-Http-Logging,Web-Http-Redirect,Web-Http-Tracing,Web-ISAPI-Filter,Web-Request-Monitor,Web-Static-Content,Web-WMI,RPC-Over-HTTP-Proxy
-Restart
The machine and exchange works good (i3 with 4gb ram, and raid 1 1tb of HD) and the organization only has 15 mailboxes. No problems with even viewer and all is correct.
Thanks again. -
We have a single exchange 2010 server running hub, cas, and mailbox as a virtual machine in VM 5. Currently on version 14.3 build 123.4 running on Server 2008 R2. Clients connecting to this server are all running Outlook 2010 with a mix of OS between
xp and win 7. Domain controllers are all 2003 sp2. We have a multi-campus environment with most users accessing across a WAN connection (mpls and metro-e). Quota Limits of 1gig imposed on all mailboxes. We do still
have a Exchange 2003 server that we are completing the migration of users off of in our environment. But all mailboxes on exchange 2010 are accessed thru Outlook 2010. Migration from 2003 to 2010 happens only after Office 2010 is installed.
We are a little over 50% migrated - (~750 users, 1100 entities)
I have some 2010 clients complaining that they receive "the server is unavailable" when launching outlook for the first time in the morning. A second time it usually opens. This happens for remote WAN and local LAN users. Also,
during the day as outlook is open, when trying to open an email or navigate between folders in the inbox they will also have outlook stop responding as it tries to contact the server.
Monitoring the server - CPU averages 40-50% thru the day with spikes in the 75-100%. Usually the processes that spike are Store.exe or msftefd.exe.
I have verified that IPV6 is disabled on the exchange server - and it doesn't exist on the 2003 DC's. Searching for answers keeps turning up results for OOF or Outlook 2003 RPC. Neither are the case here.
Would we benefit from adding a 2nd server into our environment? If so - hub? Cas? which would be better to offload off the primary?
or is there a better way to handle this issue?Hi,
According to your description, the issue happens on every morning. Thus, it’s mostly likely caused by that there are too many sessions bringing the performance issue.
To confirm it, we can use performance monitor to check the following counters when Outlook works slowly:
RPC request, RPC average latency and RPC operation/sec
http://technet.microsoft.com/en-us/library/dd335215(v=exchg.141).aspx
If you have any question, please feel free to let me know.
thanks,
Angela Shi
TechNet Community Support -
My
environment is as follows.<o:p></o:p>
We have 3 exchange 2010 servers.<o:p></o:p>
EX1, EX2, DREX1. All servers are exchange 2010 14.2 build 247.5<o:p></o:p>
EX1 and EX2 have two nics, one LAN and one for BACKUP <o:p></o:p>
The dag is configured to to communicate to all three.<o:p></o:p>
The IP for EX1 is 192.168.100.51 - LAN, switch A; 160.100.3.51 - BACKUP, switch B<o:p></o:p>
The IP for EX2 is 192.168.100.52 - LAN, switch A; 160.100.3.52 - BACKUP, switch B<o:p></o:p>
DREX1 is located in another network. It is in another building connected via vpn. The ip
for DREX1 is 192.170.100.51. There is no BACKUP network for this server<o:p></o:p>
(These are not the actual address, but the concept is similar)<o:p></o:p>
The dag network configuration is as follows.<o:p></o:p>
DAGNetworkDR has a subnet of 192.170.100.0/24; Network Interface of 192.170.100.51<o:p></o:p>
DAGNetwork01 (LAN) has a subnet of 192.168.100.0/24; Network Interfaces of 192.168.100.51
and 192.168.100.52<o:p></o:p>
DAGNetwork02 (BACKUP) has a subnet of 160.100.3.0/24; Network Interfaces of
160.100.3.51 and 160.100.3.52<o:p></o:p>
Last Friday, 3/7/14 something happened.<o:p></o:p>
To lead up to that, I noticed the issue because we have DPM 2010 making 15minute express
full backups of our exchange databases when I reviewed the backup jobs, the
exchange database backups using dpm2010 failed all weekend long starting at
11:00pm that Friday.<o:p></o:p>
The issue that occurred was that the database copies went to a status of Disconnected and
Resynchronizing, and after about 30 seconds, the status would show
Resynchronizing.<o:p></o:p>
All attempts to resolve this with the research methods failed. Deleting the
logs and reseeding failed, suspending, then rebuilding failed.<o:p></o:p>
Creating a new database worked, creating a database copy worked, but seeding the copy
failed. All with the same pattern, a status of Disconnected and
Resynchronizing, and after about 30 seconds, and the status would show
Resynchronizing.<o:p></o:p>
The database was mounted on EX2, and seeding to EX1
DREX1 has no issues.<o:p></o:p>
All the DAG networks were in a good status, all Green and all up. I could ping from
EX1 to EX2 and from EX2 to EX1 on both networks fine.<o:p></o:p>
In the middle of the week, about 3/12/14, we resolved the issue. We rebooted EX1, no
fix, but after rebooting EX2, it all resolved. It did a fail over to DREX1. I think it failed to do the failover to EX1. However, when it came back up, it was happy, it was reseeding slowly. After about an hour and a half, it
was all healthy.<o:p></o:p>
We thought it was all good EXCEPT, on Friday night at 11:00pmish on 3/14/14, the same
issue occurred; a status of Disconnected and Resynchronizing, and after about
30 seconds, the status would show Resynchronizing.<o:p></o:p>
The database was mounted on EX2 and seeding to EX1 again. I did not notice the
issue until I performed my review of the backup jobs again on Monday.<o:p></o:p>
Late Monday morning, my boss and I were looking at more articles, but all the same results,
delete the database copy and reseed. Which we knew would not work.<o:p></o:p>
I mentioned that it was strange because the networks were all up. What made us try the next
scenarios was just "luck" I guess. I made a statement about not remembering
which networks DAG replication took place over. Again, we have a LAN and
a BACKUP network. Bother were listed to use replication, so we decided to
disable replication over the BACKUP network in the EMC. Immediately, the
database copies switch to healthy and started seeding.<o:p></o:p>
Later that evening I enabled replication over the BACKUP network and it failed again with
a status of Disconnected and Resynchronizing, and after about 30 seconds, the
status would show Resynchronizing. I disabled the replication over the
BACKUP network and the database copy went back to healthy<o:p></o:p>
The last round of windows updates were pushed to these boxes the Friday before the
issues began, 2/28/14. I don't believe the updates to be the issue. <o:p></o:p>
Can anyone make any suggestions as to where to look? It worked with the BACKUP
network enabled to so long that is eluding me as to why the issues would start
all of a sudden. I am also confident that the switch configuration for the
BACKUP network has not been changed at all in at least 2-3 months. I am
still open to look at the config. Please help if you can.<o:p></o:p>
<o:p></o:p>
Lastly, this is more of a monitoring thing, if anyone can suggest a good way to monitor
the database copies so that I can be alerted when the status changes. We
use WhatsUpGold 16.2 to monitor the devices. <o:p></o:p>
Thanks for any future assistance.<o:p></o:p>
Ian
IanYou don't have to enable the Replication on Backup Network.
You should collapse your DAG network.
http://blogs.technet.com/b/timmcmic/archive/2011/09/26/exchange-2010-collapsing-dag-networks.aspx
http://blogs.technet.com/b/samdrey/archive/2012/12/04/exchange-2010-dag-collapse-the-mapi-networks-in-a-dag.aspx
Post the result of Below command:
Get-DatabaseAvailabilityGroupNetwork
Cheers,
Gulab Prasad
Technology Consultant
Blog:
http://www.exchangeranger.com Twitter:
LinkedIn:
Check out CodeTwo’s tools for Exchange admins
Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose. -
Hi,
Exchange 2010 MP fails to run Troubleshoot-DatabaseSpace.ps1. Results in a Warning with "Error formatting a string: Format string is not supported". The EventNumber is 402. The Microsoft article it points to is useless at http://technet.microsoft.com/en-us/library/749e0eac-ebb2-41e3-8fa2-4a03a1bd3571.aspx
Run ".\Troubleshoot-DatabaseSpace.ps1 -Server MailboxServer.domain.com -MonitoringContext" and works fine.
Any help appreciated.
thanksHi,
Before the newer MP release, we can disable these 4 monitor via override and this should not run the Troubleshoot-DatabaseSpace.ps1 :
KHI: Failed to execute Troubleshoot-DatabaseSpace.ps1.
KHI: The database copy is low on database volume space and continues to grow. The volume is under 25% free
KHI: The database copy is low on database volume space and continues to grow. The volume has reached error levels under 16% free.
KHI: The database copy is low on database volume space and continues to grow. The volume has reached critical levels 8% free.
Alex Zhao
TechNet Community Support
Maybe you are looking for
-
External drive connected to FW800 ExpressCard - bootable?
I have a 2.0 GHz Core [1] Duo MacBook Pro with no built-in Firewire 800, and wish to use an external boot drive at faster-than-FW400 speeds. Trying eSATA seems to be an endless nightmare right now, but what about FW800? If I connected a FireWire 800
-
How do I change the name of my iTouch? The one visible when you sync in iTunes
How do I change the name of my iTouch ? The one visible when you sync the device in iTunes, same as for iPhones, etc
-
A symbol appears in lower right corner of screen???
I get this symbol appearing in the lower right of my screen on my w510. It is a capital A with what looks like a "house" around it. Can anyone tell me what it is and how to get rid of it? Tom Solved! Go to Solution.
-
Sunstudio 11 - dbx crashes on c++ with some special stdio flags
Hello, when I compile the following C++ code (separate header and implementation file) with "g++ -gdwarf-2" or just "g++ -g": - CClass.h: #ifndef CClassH #define CClassH class CClass public: CClass(int _i); ~CClass(); private: int i; #endif /* CClass
-
Creating a second version series for the family
If i have already create a family say "Family A" and the version series say "Version Series A" under Family A. What are the step to create a second version series say "Version Series B" under Family A using IFS API