Insufficient system resources exist
Cecelia
added space to a server. After I did that, DPM required a consistency check
which it finished successfully. Ever since then I am unable to do a
synchronization. I can do consistency checks then do a recovery point but
synch's fail everytime. All of my other protection groups are synching
successfully. The error I receive is: "An unexpected error occurred during
job execution. (ID 104 Details: Insufficient system resources exist to
complete the requested service (0x800705AA))." I did try changing the
pagefile from letting the O/S (Windows 2008) manage it to assigning it the
recommended amount. But when I did that all of my DPM jobs started failing.
So I changed it back to managed but left the amount, then every was fine
again except for my original issue with that one protection group's synch
failing. Any ideas?
Hazem.Bahgat
modify the disk location on the DPM protection group, increase the size a
little bit
Cecelia
of it. But I added just a little bit more, 20GB, and I still recieve the
same error which synchronizing.
Saurabh Bansal [MSFT]
GENERAL RECOMMENDATIONS
============================
1. DPM and Memory
As per http://technet.microsoft.com/en-us/library/bb795826(TechNet.10).aspx
DPM requires a pagefile size that is 0.2 percent the size of all recovery
point volumes combined, in addition to the recommended size (generally, 1.5
times the amount of RAM on the computer). For example, if the recovery point
volumes on a DPM server total 3 TB, you should increase the pagefile size by
6 GB.
NOTE: There is a Volume Shadow Copy Service (VSS) non-paged pool limitation
on x86 32-bit operating systems. Therefore, if you are protecting more than
10 TB of data, the DPM server must be running on a 64-bit operating system.
Note: The problem can occur on the protected systems when not enough
non-paged memory available to the kernel components mainly Dpm VolumeFilter,
due to 32 bit architecture limitation.
On a 32 bit system, the maximum available non-paged pool memory is 256 MB
irrespective of physical RAM size. This memory is shared by various kernel
components including VolSnap and Dpm VolumeFilter.
Dpm VolumeFilter requires 8 MB + additional contiguous 8 MB for every
TeraByte of volume.
[for 2TB this becomes 16MB + contiguous 16MB]
When the memory becomes fragmented (which easily happens after running for
sometime), the contiguous 8MB memory allocation fails causing this error
(0x800705aa).
Also the Workaround/Solutions include:
a. Use ForceFixup key as described in the workaround section below. This
removes the Dpm VolumeFilter memory requirement.
b. Go with a 64 bit system. - This will solve the problem, but is a very
hard requirement to change the system
c. Go with Smaller volumes - This relaxes the contiguous memory requirement
and reduces the frequency of failures.
2. Install this Windows 2003 fix to address issues with Memory Manager
trimming operation.
953600 Message after multiple unsuccessful calls to certain functions on a
computer that is running Windows Server 2003 or Windows XP Professional x64
Edition: "Running out of physical memory"
http://support.microsoft.com/default.aspx?scid=kb;EN-US;953600
NOTE: The above fix contains ntoskrnl.exe version 5.2.3790.4301 and INCLUDES
the two fixes listed below.
938486 A Windows Server 2003-based computer becomes unresponsive because of
a memory manager trimming operation that is caused by an indeterminate
module that requests lots of memory
http://support.microsoft.com/default.aspx?scid=kb;EN-US;938486
942486 The System process may consume most of the available system memory on
a multiprocessor computer that is running Windows Server 2003
http://support.microsoft.com/default.aspx?scid=kb;EN-US;942486
3. Add/Modify the registry settings below to trim the Paged Pool usage of
the server.
[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Memory
Management]
"PoolUsageMaximum"=dword:0x0000001E (This is 30 decimal)
"PagedPoolSize"=dword:0x0d800000
See this blog:
http://blogs.technet.com/perfguru/archive/2008/06/24/how-to-troubleshoot-event-id-333.aspx
dword value meanings:
0x11800000 (which gives you about 280 MB)
0xC0000000 (which gives you about 250 MB)
0x0d800000 (which gives you about 216 MB)
Note Setting the PagedPoolSize value to 0xFFFFFFFF allocates the maximum
paged pool
to the computer but is ignored if /3gb is in the boot.ini file.
4. 948496 An update to turn off default SNP features is available for
Windows Server 2003-based and Small Business Server 2003-based computers
http://support.microsoft.com/default.aspx?scid=kb;EN-US;948496
5. Change the Video drivers to generic VGA driver - that can give back up to
10MB memory.
6. Disable NIC teaming and update NIC drivers (if HP server, use the latest
support pack to update driver)
7. Install the Hotfix below to update STORPORT.SYS to the latest version.
946448 You cannot remove a persistent iSCSI target if Physical Address
Extension (PAE) is enabled on a Windows Server 2003-based computer
http://support.microsoft.com/default.aspx?scid=kb;EN-US;946448
Note: The above hotfix includes STORPORT.SYS version 5.2.3790.4206)
EXCHANGE 2003 SPECIFIC RECOMMENDATIONS
==========================================
1. Decrease the Userva value in the Boot.ini from 3030 to 2970 as shown
below.
NOTE: If the problem is with non-pagd pool - then modifying this will not
help, however if the problem is with paged pool memory or available PTE's ,
you can decrease this setting all the way down to 2900 and still be
supported by exchange, but start with 2970.
[boot loader]
timeout=30
default=multi(0)disk(0)rdisk(0)partition(1)\WINDOWS
[operating systems]
multi(0)disk(0)rdisk(0)partition(1)\WINDOWS="Windows Server 2003, Standard"
/noexecute=optout /fastdetect /3GB /Userva=2970
2. When a system only has 128MB max for Non-Paged pool memory even small
leaks like the AfdP leak which may only leak 1MB a week can be enough to put
an exchange server over the edge, especially given that often times they are
not rebooted for some time. So install the following Hotfixes below to
update AFD.SYS and
SMTPSVC.DLL.
939180 Stop error message when an application that uses Winsock is running
on a Windows Server 2003-based computer: "Stop 0x0000007E"
http://support.microsoft.com/default.aspx?scid=kb;EN-US;939180
Note: The above hotfix includes afd.sys version 5.2.3790.4106.
937753 Messages that are sent from multiple domains are not delivered to
Exchange Server 2003 installations that are running on Windows Server
2003-based computers
http://support.microsoft.com/default.aspx?scid=kb;EN-US;937753
Note: The above hotfix includes smtpsvc.dll version 6.0.3790.4125
3. Exchange Server 2003 Service Pack 3 Hotfix (SP3 meaning it is scheduled
to be included in SP3)
908674 Stop error message on an Exchange Server 2003 server:
"INVALID_PROCESS_ATTACH_ATTEMPT (5)"
http://support.microsoft.com/default.aspx?scid=kb;EN-US;908674
4. You can add the following registry entry to allow HTTP to use more
Non-paged pool memory before it stops servicing requests. HTTP is designed
to stop servicing requests at Total Non-Paged Pool minus 20MB, unfortunately
on a system running /3GB that often does not work well:
HKLM\SYSTEM\CurrentControlSet\Services\HTTP\Parameters
EnableAggressiveMemoryUsage=1 (DWORD)
5. If running Exchange 2003 SP2. The exchange EXIFS (AuxL and Flst tags) can
be moved from Non-paged Pool back to Paged Pool (Behavior that existed prior
to Exchange 2003 SP1)
HKLM\SYSTEM\CurrentControlSet\Services\EXIFS\Parameters
AuxFreeListInPagedPool=1 (DWORD)
FreeListInPagedPool=1 (DWORD)
WORKAROUND
==============
IF all of the above recommendations do not help resolve the issue, you can
try enabling this registry key on the protected server to possibly bypass
the problem with non-paged memory resources, but it will effect DPM recovery
point and consistency check performance on the protected server. Use this
workaround as a last resort option only.
A) On the protected server, browse to the registry path
HKLM\Software\Microsoft\Microsoft Data Protection Manager\Agents\2.0.
B) Create a DWORD registry value called "ForceFixup" and set it to 0x1.
C) Restart the DPMRA service on the protected server.
D) On the DPM server, re-run the failed consistency check, synchronization,
or
recovery point job.
NOTE: Recovery point and CC may take up to 3 times longer (or more) to
finish when forcefixup key is added, so set customer expectations
accordingly.
Thanks
Saurabh
"Cecelia" <Cec...@discussions.microsoft.com> wrote in message
news:9EE1FF88-142A-4F99...@microsoft.com...
Fregoso
issue.