NBU 5.1 MP3

[Ricardo_Wagemak]

Hello,

I am at a currently working on a Windows Server 2003.

System SPEC

OS NameMicrosoft(R) Windows(R) Server 2003, Standard Edition
Version5.2.3790 Service Pack 1 Build 3790
OS ManufacturerMicrosoft Corporation
System ManufacturerDell Computer Corporation
System ModelPowerEdge 2850
System TypeX86-based PC
Processorx86 Family 15 Model 4 Stepping 1 GenuineIntel ~2793 Mhz
Processorx86 Family 15 Model 4 Stepping 1 GenuineIntel ~2793 Mhz
Processorx86 Family 15 Model 4 Stepping 1 GenuineIntel ~2793 Mhz
Processorx86 Family 15 Model 4 Stepping 1 GenuineIntel ~2793 Mhz
BIOS Version/DateDell Computer Corporation A01, 2-9-2004
SMBIOS Version2.3
Windows DirectoryC:\WINDOWS
System DirectoryC:\WINDOWS\system32
Boot Device\Device\HarddiskVolume1
Hardware Abstraction LayerVersion = "5.2.3790.1830
Total Physical Memory2.047,27 MB
Available Physical Memory1,05 GB
Total Virtual Memory3,85 GB
Available Virtual Memory2,81 GB
Page File Space2,00 GB

Running Veritas Netbackup 5.1 MP3 and although everything works it is taking too long to start writting to tape

In job monitor
(Operation) it states writting.
(Detailed Status) it states positioned.
(Kilobytes) it states 0

On the netbackup client icon, current status.

Actual states all (0)
Estimated (lots of files, and kilobytes) and continue to increase.

The actual data does not start to be written to tape for about 2 hours.

Is there something I need to change or do, so that it can start sendding data to tape directly after I start the backup.

It just seems to take too long to calculate the amount of files to be backed up.

I have LTO2 drives and SAN attached Volumes.

Many Thanks

Ricardo

[zippy]

Ricardo,

1: How mutch data are you backing up?
2: how many images are on the tape your using for the backup?
3: Do you have MPX set on this backup?
4: Do you have compression set?
5: Are you using a Fibre SCSI bridge?
6: What are the settings in within the Bridge?
7: What speed (1gig - 2 gig - 4 gig) do you have you switch port set to, check between the server to the switch then to the bridge to the tape drive.
8: Does the winders server backup up over the LAN to the Master or SSO server then to tape? If it does have you checked the speed setting on the Winders LAN card and the Switch to make sure all have the same speed?

2 hours may not be all that bad, of course it depends on all the questions above.

JD

[Ricardo_Wagemak]

Hi James,

1 - E:\ 1.5TB, F:\1.5TB, G:\2.0TB (1 full & 5 inc per week.)
2 - New tapes every day.
3 - MPX x8 on 2 tape drives, (4 streams per drive)
4 - Hardware compression
5 - I have to find out.
6 - I have to find out.
7 - 2 x Qlogic 2GIG 1 for Tape drive and 1 for disc drives.
8 - Local Master/Media/Client (SAN tape drives) (NO Network involved)

I will find out more about the SAN layout of this system and post it as soon as possible.

Many Thanks

Ricardo

[Dennis_Strom]

How are you backing up your file systems. backing up 1.6T in one stream to me is a large chunk. break up each file sytem into a few streams.
NEW_STREAM
file list
NEW_STREAM
file list
.....

MPX8 may be a bit aggressive.

[Ankur_Kumar]

Please do look into the buffer settings given below

Modification:
For most configurations, the default NetBackup buffer settings are correct and there is no need to adjust them for the purpose of performance. Furthermore, there are factors outside of NetBackup which effect performance and should be reviewed. Some of these external factors include Host Bus Adapter (HBA) cards, SCSI cards, network interface card (NIC) settings, client disk I/O speed, network latency, and tape drive I/O. All of these should be reviewed to determine their respective impact on backup and restore speeds prior to any attempts to tune NetBackup.

On a Windows server, four different buffer settings can be modified to enhance backup performance. Those settings are:
�NUMBER_DATA_BUFFERS: The number of buffers used by NetBackup to buffer data prior to sending it to the tape drives. The default value is 16.
�SIZE_DATA_BUFFERS: The size of each buffer setup multiplied by the NUMBER_DATA_BUFFERS value. The default value is 65536.
�NET_BUFFER_SZ: The network receive buffer on the media server. It receives data from the client. The default value is 256K.
�Buffer_size: The size of each data package sent from the client to the media server. The default value is 32K.

Overview:
When a backup is initiated, the client packages data of the amount specified by the Buffer_size value, and transfers the information to the media server, which in turn, buffers that data in the NET_BUFFER_SZ. When the NET_BUFFER_SZ is full, it transfers data to the array of space created by a combination of NUMBER_DATA_BUFFERS and SIZE_DATA_BUFFERS. As soon as at least one of the SIZE_DATA_BUFFERS is full, assuming the drive is ready to write, the information is written to the tape drive.


To troubleshoot performance issues related to buffer settings, enable and review the bptm log on the media server and the bpbkar log on the client. On the media server, go to the \Veritas\NetBackup\Logs directory, and create a bptm folder. On the client, go to the
Examine the bptm and bpbkar logs for references to waits and buffers and compare one side with the other. Individually, each side's number of waits means nothing. Only when compared to the opposite side, can you deduce where a potential bottleneck is:


Warmest Regards
Ankur Kumar

[Ricardo_Wagemak]

Thank you for all your advice and help. Eventually the performance issue was because they build the servers here with 1 x 1GB HBA card and all storage and tape drives are going through the same HBA.

I'll need to split that up and then things will be resolved.

Many Thanks

Ricardo

[sdo]

Hi Ricardo,

Another reason that can cause a delay to the start of Windows is the "tracker" job on the client. The tracker has three functions:
1) To provide an icon in the sys tray (quick access to console or backup and restore gui)
2) Total up files and size before a backup so that the estimate is more accurate
3) Provide a GUI to allow the client to track the backup progress, i.e. %age done.

This functionality is almost never needed, and again is another "standard" install feature that isn't really required.

In v5.x and above, the tracker isn't started via a normal "startup" shortcut, but is instead started via a rgistry key.

This document explains how to disable/remove "tracker" from starting up:
http://support.veritas.com/docs/267253


My $0.02 recommendation is to always remove tracker. If a client has many folder/files then the "pre backup count/sizing" feature can cause quite a delay.

The ability to track backup progress from the client side is almost never needed, as admins can quite easily track backup progess from the admin console.

Regards,
Dave.