Veritas-bu

[Veritas-bu] backup took longer for LTO drive

2003-04-11 08:25:38
Subject: [Veritas-bu] backup took longer for LTO drive
From: dbase77 AT yahoo DOT com (Feroz F. Basir)
Date: Fri, 11 Apr 2003 13:25:38 +0100 (BST)
Hi,

WOW, that is very fast 15-20 MB/s transfer rate. I've
check the bptm log file. It indicated 3.5 MB/s speed.
That is very slow. Do you mind share your
configuration, please? How many multiplex? Did you
allow multi volume? What is your buffer size and no.
of buffer? I want to use only 1 tape drive here.

If I use 2 tape drives at the same time, with 2 policy
and split filesystem between 2 drives it took only
less than 1 hour which is very good. This is weird.
Am I missing something here you think? Im using SUN
storedge L25 (tape library) and D2 (External disk with
veritas volume manager).

regards,
feroz

 --- Sergey Tsyganenko <tsyganenko AT vectis DOT ru> wrote: >
On Fri, 2003-04-11 at 14:14, Feroz F. Basir wrote:
> > Hi,
> > 
> > I've been running a few tests for our backup on
> > solaris 8 sparc with netbackup businesserver 4.5
> GA.
> > The backup took about 7 hours for 93 GB data.
> Setting
> > as follow:
> > 
> > multiple volume = NO
> > multiplex = NO
> > Buffer size = 262144
> > no. of buffer = 16
> > network buffer = 262144
> > tape library with 1 robot and 2 LTO drives.
> > 
> > I want to use ONLY 1 tape drive. What am I missing
> > here? Anybody could share their experience?
> 
> LTO is *EXTREMELY* depends on data supply speed.
> 
> Example from our real life:
> 
> Some background.
> NetBackup DataCenter 4.5GA, Solaris 7, StorEdge L40
> ( 40 tapes, 2 LTO
> drives ), 1+ TB data under vxvm/vxfs control. Sun
> UE4500 10xUSII/400Mhz
> cpus, StorEdge L3910 - 6 x fully filled with 73Gb
> drives T3+. Server is
> under ubnormal heavy load ( LA is about 90 during
> worktime, 50-60 during
> nighttime, run queue between 20 and 46 ), no disk io
> bottleneck.
> 
> At early stage due to bug in netbackup we forced to
> do non-multiplexed
> 1-way backup of data to one drive ( second drive
> must be free to serve
> day-to-day jobs such as redologs archivation and
> filesystems backup ).
> 1-way backup took more than 3-4 days. "iostat -xcn 5
> | grep rmt"
> indicates that data transfer to tapes was under 6-9
> mbytes per second
> and sometimes falls to 1200 kb/s and even zero.
> 
> I've found workaround of netbackup bug and we
> configured 4-way
> multiplexed backup to one drive. As for now it tooks
> 21-25 hours.
> "iostat -xcn 5 | grep rmt" shows permanent 15-20
> Mb/sec throughput.
> 
> -- 
> Sergey Tsyganenko <tsyganenko AT vectis DOT ru>
>  

__________________________________________________
Yahoo! Plus
For a better Internet experience
http://www.yahoo.co.uk/btoffer

<Prev in Thread] Current Thread [Next in Thread>