TCPWARE instead of TCP and SQL * NET compatibility

Similar Questions

• Lock on a table and SQL * Net more data from client

  Hello

  I have a problem loading data through Php in Oracle: web page of Php open an xml file, create a .csv, then begins to Oracle using an external table to process the data.

  The problem is that, the second time that PHP is trying to run the following query:

  INSERT INTO table1 T (Field1, Field2, field3, field4, sphere5, case field6) VALUES ('Aaa1E63819707', SYSDATE, 'KO', 'error', 'L', 'file_name.xml') POLL INTO Field1: id

  on the Oracle sessions, I have an exclusive lock on table1 and wait for SQL * Net more data from client.

  Where is the problem?

  Thank you

  Igor

  Finally I found the problem: a DBA changed the cursor sharing EXACT force! I restored the CORRECT setting and now everything is fine.

  Thank you

  Bye,.

  Igor

• Several alter session and SQL * Net message from dblink in APEX

  Hello

  I have Apex 4.2.3 and I use Glassfish more listener of the Apex.

  Database is 11.2.0.4 under linux

  Looking for the session active database, I see that apex run several alter session with a lot of expectation as SQL * Net message from dblink the.

  Ex:

  ALTER session set NLS_TERRITORY = "AMERICA."

  call the query of disc elapsed to cpu count current lines

  ------- ------  -------- ---------- ---------- ---------- ----------  ----------

  Parse 58 0.00 0.00 0 0 0 0

  Run 58 0.06 3.53 267 58 0 0

  Fetch        0      0.00       0.00          0          0          0           0

  ------- ------  -------- ---------- ---------- ---------- ----------  ----------

  Total 116 0.06 3.53 267 58 0 0

  Chess in the library during parsing cache: 0

  The analysis of the user id: 277 (recursive depth: 1).

  Elapsed time are waiting on the following events:

  Event waited on times max wait for the Total WHEREAS

  ----------------------------------------   Waited  ----------  ------------

  SQL * Net message to dblink 174 0.00 0.00

  SQL * Net message from dblink 174 3.47 0.03

  ********************************************************************************

  
  

  Why he is doing a lot to change session? Why the wait is SQL * Net message from dblink the?

  We have recently to load a new language, but I'm not sure that this is the main problem.

  
  

  
  

  I think that you have the link db in the workspace schema and data requested by the Apex of this link. For example, it may be

  report which uses db link. Check dba_db_links or user_db_links for the existing links. Check if this link used by Apex.

• Wonder if I do 1z0-064 12 c perf and tuning instead of 11 g sql tuning?

  Hello

  Wonder if I do 1z0-064 12 c perf and tuning instead of 11 g sql tuning?

  I spent so much time studying from 11 g sql tuning.

  Basically, I used the book of oracle for it certification preparation.

  I don't have the funds that I used to.

  All thoughts.

  I guess there is a lot of overlap between 1z0 - 117 and 1z0 - 064.

  I wonder if Oracle will never provide 20% discount for 12 c perf and tuning.

  Did anyone will ever write a certification for this book.

  Thank you Roger

  user6862024 wrote:

  Wonder if I do 1z0-064 12 c perf and tuning instead of 11 g sql tuning?

  I wonder if Oracle will never provide 20% discount for 12 c perf and tuning.

  Oracle gives RARELY given to reviews.  They can provide a discount on a package bundled (say buy a certification exam voucher and they will dismiss a practical review of 1 of 2 certified vendors: Self Kaplan or transcend.).

  user6862024 wrote:

  Did anyone will ever write a certification for this book.

  Well, the main author who is not written of Oracle Press books, is Matthew Morris who wrote study guides specifically for the Oracle Certification exams.

  His site is here--> Prep of Oracle Certification

  And there you will see that a link for any study of his guide at the top right.

  And it also provides links to study materials to review you're looking for.  --> Oracle Certification Prep: details of the examination and resources in preparation for 1Z0 - 064

  Currently, it is NOT a guide for exam 1Z0-064, but it may be trying to write a.  I don't know that Matthew will respond as it usually check this subforum.

• SQL * Net and Auditor job details

  Hello
  
  I am aware of SQL * Net and auditor of the concepts.
  
  SQL * Net Treaty of the transmission of data between the client to the server and vice versa.
  
  Listener established the connection between the user and the server process.
  
  I know that on the Oracle server, there are two. ORA files available in the file Admin/network for SQL * Net and the listener.
  
  I am interested in knowing the following things.
  
  (1) in my understanding, process user run the server and not the client.
  
  (2) Please share me the stream used between SQL * Net listener and how the flow occurs.
  
  After the TOAD or SQL * Plus creates connectivity and gets the Session, how they still work.
  I understand it, once the connection has been established, even when the listener goes down.
  SQL * over the connection to the server remains intact.
  
  Can you please share more fundamental and work know on these two areas, it is not 100% clear to me, thank you.

  See the post of Ed - http://edstevensdba.wordpress.com/2011/02/09/sqlnet_overview/

  Srini

• SQL * Net message from client

  Hi all

  I have a session that runs longer than usual. Instead of filling in 20 minutes, he was running for about 2 hours.

  Initial review of show expected that the session was waiting on this SQL * Net client message for a piece of time.

  The session state was idle, while the State changed pending work very frequently.

  I had activated track and found the bottom of the entries in the trace file.

  EXEC #11529215044976967776:c = 0, e = 1944, p = 0, cr = 11, cu = 3, set = 0, r = 0, dep = 0, og = 1, plh is 2651636916, tim = 2309884016239

  #11529215044976967776 ERROR: err = 54 tim = 2309884016520

  
  

  TKProf reports the following

  TOTALS FOR ALL NON RECURSIVE INSTRUCTIONS

  call the query of disc elapsed to cpu count current lines

  ------- ------  -------- ---------- ---------- ---------- ----------  ----------

  Parse        0      0.00       0.00          0          0          0           0

  Run the 884 1.47 1.42 2660 9724 0 0

  Fetch        0      0.00       0.00          0          0          0           0

  ------- ------  -------- ---------- ---------- ---------- ----------  ----------

  884 total 1.47 1.42 2660 9724 0 0

  I need to know what is meant by the error line in the trace file?

  We use oracle EBS 11i with 11 GR 2 db. This session belongs to a simultaneous request

  
  
  

  Oracle - ORA-00054: resource busy and acquire with NOWAIT specified or expired timeout - Stack Overflow

  -------------

  Sybrand Bakker

  Senior Oracle DBA

• SQL * Net more data from client wait event

  Hi all

  DB 11.2.0.2

  AIX 6

  I get two events of high expectation of the AWR report page

  (1) SQL * Net more data from client

  synchronization of log files 2)

  Need advice to the configuration of the hardware and network latency?

  What should I do for the first event of waiting?

  This simple average may, what they're doing massive data loads with a very large table size that the Oracle Server manages very effectively but necessarily expect much coming data of the customer and parallel large log files written to end on the validation.  It is possible that setting a large SDU and tcp large transmiat and receive buffers would help if this happens to be what they do; Similarly, if that's what they do, then reduce their size of table can also have some effect.

  Concerning

  Jonathan Lewis

• Fact ' SQL * Net client message' indicate that the network is not good?

  The report below 10046 trace, I have just issued a query to extract all records from a table with about 400 K records, but it took more than 3 minutes to recover all data.
  Most of the time when waiting for SQL * Net message from client.
  This indicates that the network is not good?
  
  
  SQL ID: d84ta7m8h6wg2
  Plan of Hash: 3635621446
  Select * from XXXXXXXXXXXXXXXXX
  
  
  call the query of disc elapsed to cpu count current lines
  Parse 1 0.00 0.00 0 0 0 0
  Run 1 0.00 0.00 0 0 0 0
  Fetch 855 0.88 0.10 0 7912 0 426721
  Total 857 0.10-0.88 0 7912 0 426721
  
  Chess in the library during parsing cache: 0
  Optimizer mode: ALL_ROWS
  The analysis of the user id: 123
  
  Rows Row Source operation
  426721 TABLE ACCESS FULL XXXXXXXXXXXXXXX (cr = 7912 pr = 0 pw = time 0 = 467817 u)
  s cost = 1847 size = map 46080540 = 415140)
  
  
  Elapsed time are waiting on the following events:
  Event waited on times expected maximum wait the Total WHEREAS
  
  SQL * Net message to client 856 0.00 0.00
  SQL * Net client message 856 3.71 179.63
  Disk file I/O 1 operations 0.00 0.00
  SQL * Net more data to the client 5224 0.37 0.46
  
  Published by: NathanielNie on January 24, 2013 16:40

  Run you the same Oracle_Home SQL and Server query as the ETL engine? (Want to use the same libraries and host and network such as ETL).

  The large data sets in ETL can be assembled with network optimization. Some of it is on the physical network layer (for example having a dedicated private network or a switched network between the ETL and the DB server), some of them can be found on the configuration settings (arraysize, configuring TCP settings to the OS layer, etc.) or libraries (for example JDBC libraries). Some of them may simply be patches to fix bugs!

  Hemant K Collette

• IPC vs SQL * Net connection

  Hi all
  
  My database is currently running on windows 2008 EE 10.2.0.4 (with all service packs). On the same machine, I have several workers (implemented as a windows service) written in c# .net using the latest drivers ODP.NET.
  To improve performance, I use PKI connection so that the TCP stack (which is used in the sqlnet connections "regularly") is shorted. I confirmed that ICP is actually currently in use:
  
  H4. Journal of the listener

  13-NOV-2009 03:01:34 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:34 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:34 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:35 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:35 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:35 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:35 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0
  13-NOV-2009 03:01:35 * (CONNECT_DATA=(SERVER=DEDICATED)(SID=orclwin)(CID=(PROGRAM=C:\Simulation\Agent.exe)(HOST=WIN2008EE)(USER=LOCAL?SERVICE))) * (ADDRESS=(PROTOCOL=ipc)(PIPENAME=\\.\pipe\EXTPROC0ipc)) * establish * orclwin * 0

  However, when I look at the awr report I see sqlnet expected on top:
  
  H4. AWR waits Albums

  Top 5 Timed Events                                         Avg %Total
  ~~~~~~~~~~~~~~~~~~                                        wait   Call
  Event                                 Waits    Time (s)   (ms)   Time Wait Class
  ------------------------------ ------------ ----------- ------ ------ ----------
  SQL*Net message to client         3,014,783   5,086,969   1687 ######    Network
  SQL*Net more data to client       1,712,644   2,441,249   1425 ######    Network
  log file sync                       695,446      75,297    108 ######     Commit
  log file parallel write             679,437      33,288     49  916.9 System I/O
  db file parallel write               49,057      22,149    452  610.1 System I/O

  I read the ticket of Taner Poder [http://blog.tanelpoder.com/2008/02/07/sqlnet-message-to-client-wait-gotcha/] about SQNet message to the customer
  and his other observations on the SQLNet more data to the client [http://blog.tanelpoder.com/2008/02/10/sqlnet-message-to-client-vs-sqlnet-more-data-to-client/]
  to get the best understanding of them.
  
  The question is why observe specific SQLNet connection is when you use IPC connections?
  
  I enjoy all the comments.
  Thank you
  Rafal.

  Oh yes - one comment more.

  So that the server process indicate if the wait is over TCP/IP or IPC, he needs to know the mechanism of transport used.

  It is very likely that the wait event is registered by the server process when it is calling its send() data to the client API. As he calls the send(), it records that the current wait state. He has no idea what the underlying code uses as a transport mechanism. In server code and client, the transport layer is one of the lower layers. So earlier in the application code, you don't know what is the actual transport (remember that Oracle also support, or has supported, other layers of communication such as Lu6.2).

  So the application layer, you deal with the transport layer as a bit of a black box. In the Oracle terminology, this is called SQL * Net layer. And may be something that implements the transport layer.

  So, this is why it is dangerous to make assumptions that pending top-level State indicates what a low level wait state is.

• Why the continuous measurement and Logging in LabVIEW 2012 Project uses strings instead of Enums and orders from the queue?

  Hello!

  I noticed that the continuous measurement and a project in LabVIEW 2012 Logging using chains instead of enums and orders from the queue. I wonder if there is a good reason for it?

  Kind regards

  Anguel

  First, string vs enum debate is probably the version of LabVIEW vim vs emacs. There are good arguments on both sides, and I doubt that there is always a "winner".

  A brief summary of our reasoning for the current state of the project examples:

  1. We used enums for the state machine because it is self-contained. A state machine will never tell himself to enter a State, he does not know. Knowing (as the programmer) all possible States with the help of an enum allows you to enlist the compiler in order to help us avoid mistakes to change the time (because you can't quite out an enum and LabVIEW can be said if you are not covering all cases to a structure of the case, etc..).

     Enums provide greater protection and rigidity by ensuring all withdrew at the time of publishing. This is often the 'default' recommendation that we do.

  2. We used strings for messages in queue manager because the producer of message and the message handler could be independent processes that are reused or traded. Channels avoid the need for the compiler to be able to connect the orders and push this responsibility to the programmer. This allows you to develop some sub-components independently as long as you agree to a series of channel commands that you can manage - you need not to share a file 'messages.ctl' or 'states.ctl '. It is conceivable a loop of message management a message it does not, how you can decide to either silently ignore it or will trigger an error (as we do in the model). The strings make it also easier if you want to swap the queues of LabVIEW outside by a TCP implementation for network vacilitate or intra-Processuse communication where the other end may or may not be written in LabVIEW.

     Channels to provide more flexibility (that is, you can add new commands to an existing via plugins system, you can pass parameters as part of the string, etc.) at the expense of pushing her potential errors at run time and to put more responsibility on the programmer.

  3. The actor's gifts frame a 3rd option - using classes such as messages. For me, it combines many of the advantages of these two enumerations (strictly typed, change errors) and strings (flexible and scalable), but with the disadvantage of being somewhat less transparent (you understand OO, be comfortable to navigate through a multitude of screws, legacy of understanding, etc.).

  I don't know there are other reasons, others to the breast OR had or seen as we validated models and examples of projects in-house, but here are my reasons. We know that we can not design for each situation there - our goal is to get useful models against new users to make them aware of what well thought LabVIEW programs are similar to experienced users know their applications better and I hope they do not hesitate to change what we provide or create their own designs, when they feel it's necessary. (On a side note, please share what you come up with - a community of experts sharing models would be really useful to us all LabVIEW users).

  Best regards

  Simon

• State SQL 28000 and SQL Server error 18452

  I have a SQL 2000 Server on a server with Windows Server 2003 computer.

  I have a laptop with Windows Vista, for which we are unable to connect to the database and it gives an error 24000 state SQL and SQL Server error 18452 with description as "failed to connect to the user (null). Reason: Not associated with approved connection SQL server.

  Please provide a solution to this problem.

  Thank you

  Hello

  Questions like these are much better handled in the TechNet IT Pro Forums.

  My moderator tools cannot transfer messages on Windows forums, please re - ask you question there.

  http://social.technet.Microsoft.com/forums/en-us/category/SQLServer/
  Jack - Microsoft MVP, Windows networking. WWW.EZLAN.NET

• Wait events, SQL * Net message from client

  Hi all

  I have a doubt about oracle wait events, specifically of "SQL * Net message from client. I've read a lot about this, I've seen some say that network and other material resources can provoke him, and saw that the problems in the application code can cause it too. I am tracking a few tests here and I don't have that chooses are the double event, usually a lot of SQL * Net message to client waiting (I used sqlplus, developer sql and pl/sql). Can someone explain to me why? Another question is where this wait type cause interference, meet it demand for the database and stays on hold and do not meet another request so that the wait is over, is it? The waiting just finish when the application closes the database connection or there is another way to end the waiting?

  Since then, thank you very much.

  Hello

  chooses double what usually causes a lot of SQL * Net message to client waiting

  Do you run them from your server applications or from the DB server itself?

  "SQL * Net client message" is an idle wait event although he'd still explain to the elapsed time of the query.

  It indicates only that the database is waiting for the other entries of the customer and when he expects the DB, those expectations get incremented.

  I suggest you explore size SDU, RECV_BUF_SIZE, and SEND_BUF_SIZE parameters.

  http://docs.Oracle.com/CD/B19306_01/network.102/b14212/performance.htm

  Kind regards

  Suntrupth

• Try to get ALUA, Multipath, Advanced TCP and NFS to export control

  This is one that I don't know that you've heard before, my script worked yesterday but does not work today. The script part, but I couldn't so he could create a CSV with all data. Today, I am the print below. 3 days ago that I have updated the remote instance of PowerCLI 4.1 to 6.3 update 1. The script was working after the upgrade.

  I would like to have the host name appear 1 a the Multipath, then the TCP and NFS info Advanced. Any help would be really appreciated

  #Use him below to select more than 1 Virtual Center to run the script against, back on back the #.

  param)

  [parameter (Mandatory = $true)]

  [string []] $vCenter

  )

   

  $esxHosts = Get-VMHost | Sort Name

  foreach ($esx in $esxHosts){

  Write-Host 'ALUA and Multipath'

  $esxcli . storage . NMP . device . list($null)

  Write-Host ' Reading TCP and NFS Advanced Configuration settings on $esx"" "".

   

  Get-VMHostAdvancedConfiguration -$vmhost $esx -name * Net.TcpipHeapSize,* Net.TcpipHeapMax,* NFS. HeartbeatMaxFailures,* NFS. HeartbeatFrequency,* NFS. HeartbeatTimeout,* NFS. MaxVolumes

  {The } Export-Csv ESX.csv NoTypeInformation - -UseCulture

  A I am right now. "

  

  Working last night but not output to CSV

  

  After spending the day working on it some more (masochist), I've gotten to the point of having the output CLOSE to there, but 4 points to solve (and 1 item from the wish list)

  1: use of the wish list $Creds automation (for inclusion in all scripts after this one)

  2: Data center to display in the name of the report for the reconciliation of the audit

  3: host name to display in the csv file,

  4: I get out of the name, the keys, the values (just name and values) on the Advanced TCP and NFS parameters

  5: am unable to get the Description of export even if I'm the see in the PowerCLI screen for $esxcli.storage.nmp.psp.list)

  # TCPandNFS_Settings - TESTS ALL TOGETHER

  [CmdletBinding()]

  (param

  [Parameter (Position = 0, mandatory = $true)]

  [System.String]

  $VirtualCenterServer,

  [Parameter (Position = 1)]

  [System.String]

  $ClusterName,

  [Parameter (Position = 2)]

  [System.Boolean]

  $asLocalUser = $true

  If ((Get - PSSnapin - name VMware.VimAutomation.Core - ErrorAction SilentlyContinue) - eq $null)

  Add-PsSnapin VMware.VimAutomation.Core

  If ($DefaultVIServers.name - no $VirtualCenterServer)

  #Determine if we are connected to VirtualCenter as a local user or if we request identification information

  If ($asLocalUser)

  SE connect-VIServer-Server $VirtualCenterServer | Out-Null

  Write-debug ' connection as a local user in the CR: $VirtualCenterServer.

  on the other

  SE connect-VIServer-Server $VirtualCenterServer - Credential (Get-Credential). Out-Null

  Write-debug ' login as a user manually selected to vc: $VirtualCenterServer.

  Write-Host "looks like we are already connected to: $VirtualCenterServer at this session.

  If ($ClusterName)

  $VMHosts = get-Cluster-name $ClusterName | Get-VMHost | Where-Object {$_.} ConnectionState - eq 'Connected'}

  $VMhosts = get-VMHost | Where-Object {$_.} ConnectionState - eq 'Connected'}

  $Report =

  foreach ($esx in (Get-VMHost))

  $esxcli = get-EsxCli - VMHost $esx

  Write-Host "reading TCP and NFS Advanced Configuration settings on $esx".

  (Get-VMHostAdvancedConfiguration - VMHost $esx - name * Net.TcpipHeapSize, * Net .TcpipHeapMax * NFS.) HeartbeatMaxFailures, * NFS. HeartbeatFrequency, * NFS. HeartbeatTimeout, * NFS. MaxVolumes). GetEnumerator

  # Select @{N = "ESX"; E = {$esx. Name}}

  Write-Host "ALUA and paths multiple parameters on $esx.

  $esxcli.storage.nmp.psp.list)

  $Report | Export-Csv "C:\Output.csv" - NoTypeInformation - UseCulture; II 'C:\Output.csv '.

  

  When the $Report is commented out export

  

• Wait for the event begins with SQL * Net message from client-time wait 178577 units

  Hello
  
  I'm watching events waiting for a request from long time in TOAD.
  I start the query on an instance of TOAD, and open the browser to log on to another instance.
  But I am surprised to see that in "TOtal expected" on the RIGHT part->
  SQL * Net message from client is the longest time and is already - > 178577 units while I just to start the query.
  
  Considering that, in the current waiting she shows DB file scattered read correctly for a few seconds.
  
  Please suggest.

  user8941550 wrote:
  Hello. No explanation for this... :-(

  Hello

  people work here, you don't know?
  I think Tom Kyte explains it well enough. This wait event is linked to your session database waiting for the guest to say to do something.

  So it is not related to the database, but to your application.
  Also as it is a wait of session event you might have had your session inactive for some time (do nothing)

  If you want to check the waiting events correctly I suggest using tkprof and start a new session in SQL more as shown by Tom Kyte in the link I posted.

  Then, run your query in sqlplus setting track and pull it out as soon as your statement is completed.
  that is to say:

  -- myest.sql
  alter session set events '10046 trace name context forever, level 12';
  SELECT ... -- your query here
  exit
  

  Run in sqlplus in this way:

  sqlplus user/password@db @mytest.sql
  

  Then check with tkprof.

  Kind regards.
  Al

• Research in migration Exchange and SQL from physical servers to virtual machines... thoughts?

  Hi all, I use VMs for several years now, however back to the use of VMWare. I am currently running ESXi 4 on 2 servers separate host, with about 16 small resource type of the applications without any problem. I heard years keep intensive stuff like Exchange and SQL out of a VM guest, for reasons of system, but now I see a ton of articles from VMWare and windows push the thing to do. I have a relatively small environment compared to what they are touting... I was wondering if there is anyone out there who has any experience/thoughts on this? If I go this route I would probably look to buy some of the nicer features instead of using just the free version of ESXi.

  My current Setup uses HP DL360G5 servers that are coming out of warranty early next year, with the disk subsystem provided by a built environment of Equallogic iSCSI, on an isolated GB network SAN. I thought rather than buy 2 replacement servers that I can put the money to the full version of vSphere with some other goodies it offers.

  My Exchange environment is a relatively small number of boxes letters, but larger in size, running 2003 currently with 200 mailboxes to about 300 GB of space. I intend to switch to 2007 or 2010 next year as part of the migration of VM. I have to go to the next version will include likely standing another server, so I might be looking to buy 2 servers instead of 1.

  My SQL environment is small to begin with, a single instance of SQL 2005 with about 20 DB 120 GB in total size.

  2 guest machines I have are two Quad-processor Xeon dual socket with 16 GB and 24 GB of RAM, with less than 50% of RAM use each... and the processing power is hardly be affected at all.

  Sorry that it's a lot of info, I hope it makes sense for someone...

  Fix.

  You just need to do some analysis of storage before you virtualize servers in order to understand the needs of e/s. But if you move storage to local disk then any loss of performance, you can get to virtualize the server is offset by the gains in terms of the solution of SAN - faster storage & drives high-performance, virtual RAID, active\active controllers, cache... etc.