Re: [Bacula-users] Data transfer rates

2006-11-15 Thread DAve
Thomas Traeger wrote:
 Excellent point. The servers that are slowest to transfer are also the 
 busiest, my front end web server and my mail gateways (Av and spam 
 filtering).

 That is certainly something to look further into. I could try turning 
 off compression for one cycle to prove the theory.

 Thanks,

 DAve

   
 You might also try reducing the compression level to gzip1/2/3 (default
 is gzip5), especially on the VLAN this might help. In my case all
 servers have a 1GBit connection to the backup system and software
 compression is useless as long as you backup on a tape drive with
 hardware compression.
 

Compression was the culprit! After much testing and trying specific 
clients and compression levels it seems that I can transfer uncompressed 
data across a 1gb network faster than I can compress the data on the 
client. This was tested last on a very busy web server, which really is 
never not busy. The overhead of compression exceeded the gains of 
transfering data compressed.

So depending on what load/cpu/ram a client has, compression should be 
altered. It is a trade off of speed verses storage space. I can adjust 
each job as required.

Just one more way Bacula's complexity pays off. Thanks everyone.

DAve

-- 
Three years now I've asked Google why they don't have a
logo change for Memorial Day. Why do they choose to do logos
for other non-international holidays, but nothing for
Veterans?

Maybe they forgot who made that choice possible.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-15 Thread Ryan Novosielski
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Did you just remove compression from the definition?

I'm still looking for a way to EXPLICITLY disable compression.

DAve wrote:
 Thomas Traeger wrote:
 Excellent point. The servers that are slowest to transfer are also the 
 busiest, my front end web server and my mail gateways (Av and spam 
 filtering).

 That is certainly something to look further into. I could try turning 
 off compression for one cycle to prove the theory.

 Thanks,

 DAve

   
 You might also try reducing the compression level to gzip1/2/3 (default
 is gzip5), especially on the VLAN this might help. In my case all
 servers have a 1GBit connection to the backup system and software
 compression is useless as long as you backup on a tape drive with
 hardware compression.

 
 Compression was the culprit! After much testing and trying specific 
 clients and compression levels it seems that I can transfer uncompressed 
 data across a 1gb network faster than I can compress the data on the 
 client. This was tested last on a very busy web server, which really is 
 never not busy. The overhead of compression exceeded the gains of 
 transfering data compressed.
 
 So depending on what load/cpu/ram a client has, compression should be 
 altered. It is a trade off of speed verses storage space. I can adjust 
 each job as required.
 
 Just one more way Bacula's complexity pays off. Thanks everyone.
 
 DAve
 

- --
  _  _ _  _ ___  _  _  _
 |Y#| |  | |\/| |  \ |\ |  | |Ryan Novosielski - Systems Programmer III
 |$| |__| |  | |__/ | \| _| |[EMAIL PROTECTED] - 973/972.0922 (2-0922)
 \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.5 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFW0fsmb+gadEcsb4RAis2AKCR+ZuQTIZ2Z8Dt3Rs7JRqCKJefnwCfchWm
WHsvjSflzU6WtA46T7PrNpY=
=8YrC
-END PGP SIGNATURE-


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-15 Thread DAve
Ryan Novosielski wrote:
 -BEGIN PGP SIGNED MESSAGE-
 Hash: SHA1
 
 Did you just remove compression from the definition?

I simply commented out the compression option in the FileSet section.

### FILESETS
FileSet {
   Name = Web6
   Include {
   Options {
 signature = MD5
#compression=GZIP
}
   File = /data/data
   File = /data/cgi-bin
   File = /data/cgivirt
   File = /data/webalizer
   File = /usr/local/scripts
   File = /data/movies
   File = /usr/local/etc
   File = /usr/local/db/mysql
   File = /etc
   }

Exclude {
 File = /etc/dnscache/log/main
 File = /tmp
 File = /.snap
 }
}

The next report stated no compression.

   FD Files Written:   1,501
   SD Files Written:   1,501
   FD Bytes Written:   2,699,586,875
   SD Bytes Written:   2,699,798,983
   Rate:   12978.8 KB/s
   Software Compression:   None

Easy.

DAve




 
 I'm still looking for a way to EXPLICITLY disable compression.
 
 DAve wrote:
 Thomas Traeger wrote:
 Excellent point. The servers that are slowest to transfer are also the 
 busiest, my front end web server and my mail gateways (Av and spam 
 filtering).

 That is certainly something to look further into. I could try turning 
 off compression for one cycle to prove the theory.

 Thanks,

 DAve

   
 You might also try reducing the compression level to gzip1/2/3 (default
 is gzip5), especially on the VLAN this might help. In my case all
 servers have a 1GBit connection to the backup system and software
 compression is useless as long as you backup on a tape drive with
 hardware compression.

 Compression was the culprit! After much testing and trying specific 
 clients and compression levels it seems that I can transfer uncompressed 
 data across a 1gb network faster than I can compress the data on the 
 client. This was tested last on a very busy web server, which really is 
 never not busy. The overhead of compression exceeded the gains of 
 transfering data compressed.

 So depending on what load/cpu/ram a client has, compression should be 
 altered. It is a trade off of speed verses storage space. I can adjust 
 each job as required.

 Just one more way Bacula's complexity pays off. Thanks everyone.

 DAve

-- 
Three years now I've asked Google why they don't have a
logo change for Memorial Day. Why do they choose to do logos
for other non-international holidays, but nothing for
Veterans?

Maybe they forgot who made that choice possible.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT  business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-11 Thread Thomas Traeger

 Excellent point. The servers that are slowest to transfer are also the 
 busiest, my front end web server and my mail gateways (Av and spam 
 filtering).

 That is certainly something to look further into. I could try turning 
 off compression for one cycle to prove the theory.

 Thanks,

 DAve

   
You might also try reducing the compression level to gzip1/2/3 (default
is gzip5), especially on the VLAN this might help. In my case all
servers have a 1GBit connection to the backup system and software
compression is useless as long as you backup on a tape drive with
hardware compression.

Thomas

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


[Bacula-users] Data transfer rates

2006-11-10 Thread DAve
Good morning,

I have a question concerning reported data transfer rates between the 
clients and the storage servers. All the above servers are connected to 
my NOC clients via a 1gb network. The servers are connected to several 
other clients via 100mb network and a 13mb VLAN.

Data rates for backups can vary widely. I have all clients running 
compression and the backup reports show compression results, varies of 
course based on the data type. I also have concurrency set to four, 
which is working.

Some examples,

1gb network
=
   FD Bytes Written:   215,582,592
   SD Bytes Written:   215,582,742
   Rate:   2245.7 KB/s
   Software Compression:   62.0 %

   FD Bytes Written:   510,861,802
   SD Bytes Written:   511,074,097
   Rate:   577.2 KB/s
   Software Compression:   80.0 %

   FD Bytes Written:   77,243,969
   SD Bytes Written:   78,145,766
   Rate:   351.1 KB/s
   Software Compression:   84.4 %

100mb network
=
   FD Bytes Written:   3,486,750,378
   SD Bytes Written:   3,488,784,689
   Rate:   2391.5 KB/s
   Software Compression:   42.4 %

   FD Bytes Written:   3,486,750,378
   SD Bytes Written:   3,488,784,689
   Rate:   2391.5 KB/s
   Software Compression:   42.4 %

13mb VLAN
=
   FD Bytes Written:   35,879,089,133
   SD Bytes Written:   35,910,549,602
   Rate:   1385.2 KB/s
   Software Compression:   41.1 %

   FD Bytes Written:   9,803,307,103
   SD Bytes Written:   9,809,416,473
   Rate:   1273.0 KB/s
   Software Compression:   57.3 %

   FD Bytes Written:   434,184,031
   SD Bytes Written:   434,204,782
   Rate:   943.9 KB/s
   Software Compression:   82.0 %


I'm not sure how to interpret the variation in rates. I don't think I 
have a problem, though I would expect the backups across the 1gb network 
to be far faster than they are. The backups across the 1gb network are 
all FreeBSD boxes, the backups across the 100mb and the 13mb VLAN are 
Windows and RedHat servers. The dir, and the sd clients are running on 
FreeBSD servers dedicated to that task, there are no other processes 
running.

Any ideas where I might start looking?

DAve


-- 
Three years now I've asked Google why they don't have a
logo change for Memorial Day. Why do they choose to do logos
for other non-international holidays, but nothing for
Veterans?

Maybe they forgot who made that choice possible.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-10 Thread John Drescher
On 11/10/06, DAve [EMAIL PROTECTED] wrote:
Good morning,I have a question concerning reported data transfer rates between theclients and the storage servers. All the above servers are connected tomy NOC clients via a 1gb network. The servers are connected to several
other clients via 100mb network and a 13mb VLAN.Data rates for backups can vary widely. I have all clients runningcompression and the backup reports show compression results, varies ofcourse based on the data type. I also have concurrency set to four,
which is working.Some examples,1gb network= FD Bytes Written: 215,582,592 SD Bytes Written: 215,582,742 Rate: 2245.7 KB/s
 Software Compression: 62.0 % FD Bytes Written: 510,861,802 SD Bytes Written: 511,074,097 Rate: 577.2 KB/s Software Compression: 80.0 % FD Bytes Written: 77,243,969
 SD Bytes Written: 78,145,766 Rate: 351.1 KB/s Software Compression: 84.4 %100mb network= FD Bytes Written: 3,486,750,378
 SD Bytes Written: 3,488,784,689 Rate: 2391.5 KB/s Software Compression: 42.4 % FD Bytes Written: 3,486,750,378 SD Bytes Written: 3,488,784,689 Rate: 
2391.5 KB/s Software Compression: 42.4 %13mb VLAN= FD Bytes Written: 35,879,089,133 SD Bytes Written: 35,910,549,602 Rate: 1385.2
 KB/s Software Compression: 41.1 % FD Bytes Written: 9,803,307,103 SD Bytes Written: 9,809,416,473 Rate: 1273.0 KB/s Software Compression: 57.3 %
 FD Bytes Written: 434,184,031 SD Bytes Written: 434,204,782 Rate: 943.9 KB/s Software Compression: 82.0 %I'm not sure how to interpret the variation in rates. I don't think I
have a problem, though I would expect the backups across the 1gb networkto be far faster than they are. The backups across the 1gb network areall FreeBSD boxes, the backups across the 100mb and the 13mb VLAN are
Windows and RedHat servers. The dir, and the sd clients are running onFreeBSD servers dedicated to that task, there are no other processesrunning.Any ideas where I might start looking?
I believe your biggest bottleneck is not the networks speed but the software compression speed as this only a few MB/s max even on the fastest pc you can get. Tape drives are able to compress data at very high speeds because they have specialized hardware (not just a cpu and memory) for that purpose.
John
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-10 Thread DAve
John Drescher wrote:
 On 11/10/06, DAve [EMAIL PROTECTED] wrote:

 Good morning,

 I have a question concerning reported data transfer rates between the
 clients and the storage servers. All the above servers are connected to
 my NOC clients via a 1gb network. The servers are connected to several
 other clients via 100mb network and a 13mb VLAN.

 Data rates for backups can vary widely. I have all clients running
 compression and the backup reports show compression results, varies of
 course based on the data type. I also have concurrency set to four,
 which is working.

 Some examples,

 1gb network
 =
FD Bytes Written:   215,582,592
SD Bytes Written:   215,582,742
Rate:   2245.7 KB/s
Software Compression:   62.0 %

FD Bytes Written:   510,861,802
SD Bytes Written:   511,074,097
Rate:   577.2 KB/s
Software Compression:   80.0 %

FD Bytes Written:   77,243,969
SD Bytes Written:   78,145,766
Rate:   351.1 KB/s
Software Compression:   84.4 %

 100mb network
 =
FD Bytes Written:   3,486,750,378
SD Bytes Written:   3,488,784,689
Rate:   2391.5 KB/s
Software Compression:   42.4 %

FD Bytes Written:   3,486,750,378
SD Bytes Written:   3,488,784,689
Rate:   2391.5 KB/s
Software Compression:   42.4 %

 13mb VLAN
 =
FD Bytes Written:   35,879,089,133
SD Bytes Written:   35,910,549,602
Rate:   1385.2 KB/s
Software Compression:   41.1 %

FD Bytes Written:   9,803,307,103
SD Bytes Written:   9,809,416,473
Rate:   1273.0 KB/s
Software Compression:   57.3 %

FD Bytes Written:   434,184,031
SD Bytes Written:   434,204,782
Rate:   943.9 KB/s
Software Compression:   82.0 %


 I'm not sure how to interpret the variation in rates. I don't think I
 have a problem, though I would expect the backups across the 1gb network
 to be far faster than they are. The backups across the 1gb network are
 all FreeBSD boxes, the backups across the 100mb and the 13mb VLAN are
 Windows and RedHat servers. The dir, and the sd clients are running on
 FreeBSD servers dedicated to that task, there are no other processes
 running.

 Any ideas where I might start looking?

 
 I believe your biggest bottleneck is not the networks speed but the 
 software
 compression speed as this only a few MB/s max even on the fastest pc you 
 can
 get. Tape drives are able to compress data at very high speeds because they
 have specialized hardware (not just a cpu and memory) for that purpose.
 
 John

Excellent point. The servers that are slowest to transfer are also the 
busiest, my front end web server and my mail gateways (Av and spam 
filtering).

That is certainly something to look further into. I could try turning 
off compression for one cycle to prove the theory.

Thanks,

DAve

-- 
Three years now I've asked Google why they don't have a
logo change for Memorial Day. Why do they choose to do logos
for other non-international holidays, but nothing for
Veterans?

Maybe they forgot who made that choice possible.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-10 Thread Ryan Novosielski
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

DAve wrote:
 Good morning,
 I'm not sure how to interpret the variation in rates. I don't think I 
 have a problem, though I would expect the backups across the 1gb network 
 to be far faster than they are. The backups across the 1gb network are 
 all FreeBSD boxes, the backups across the 100mb and the 13mb VLAN are 
 Windows and RedHat servers. The dir, and the sd clients are running on 
 FreeBSD servers dedicated to that task, there are no other processes 
 running.
 
 Any ideas where I might start looking?

There are a couple of things that jump out at me. Compression seems to
be affecting your rates a certain amount -- that probably makes sense,
as it takes some time to do compression. If you can, I'd be curious to
see what happens if you turn compression off and try the same numbers.

That said, your speed issues might be due to A) spooling or B)
misconfigured duplex settings on your machine.

Check netstat -ian or your platform's equivalent to make sure your
network settings are OK. As far as the spooling goes, you could try
without to see what happens (if you are using it), but basically my take
on that effect was that's the way it goes -- you spool and your rate
drops. I'm not thrilled that I no longer have a good way to check the
SD to tape data transfer rate, but as long as the backups run in an
acceptable amount of time, I stay happy.

- --
  _  _ _  _ ___  _  _  _
 |Y#| |  | |\/| |  \ |\ |  | |Ryan Novosielski - Systems Programmer III
 |$| |__| |  | |__/ | \| _| |[EMAIL PROTECTED] - 973/972.0922 (2-0922)
 \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.5 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFVLDxmb+gadEcsb4RAhpWAJ0cG8KPP5SmnY8XWyGv80OU8CDCuACeK0zn
PuKB65lC6TaN7Akq0Cf/XtI=
=SsjZ
-END PGP SIGNATURE-


-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-10 Thread DAve
pedro moreno wrote:
 On 11/10/06, Ryan Novosielski [EMAIL PROTECTED] wrote:

 -BEGIN PGP SIGNED MESSAGE-
 Hash: SHA1

 DAve wrote:
  Good morning,
  I'm not sure how to interpret the variation in rates. I don't think I
  have a problem, though I would expect the backups across the 1gb 
 network
  to be far faster than they are. The backups across the 1gb network are
  all FreeBSD boxes, the backups across the 100mb and the 13mb VLAN are
  Windows and RedHat servers. The dir, and the sd clients are running on
  FreeBSD servers dedicated to that task, there are no other processes
  running.
 
  Any ideas where I might start looking?

 There are a couple of things that jump out at me. Compression seems to
 be affecting your rates a certain amount -- that probably makes sense,
 as it takes some time to do compression. If you can, I'd be curious to
 see what happens if you turn compression off and try the same numbers.

 That said, your speed issues might be due to A) spooling or B)
 misconfigured duplex settings on your machine.

 Check netstat -ian or your platform's equivalent to make sure your
 network settings are OK. As far as the spooling goes, you could try
 without to see what happens (if you are using it), but basically my take
 on that effect was that's the way it goes -- you spool and your rate
 drops. I'm not thrilled that I no longer have a good way to check the
 SD to tape data transfer rate, but as long as the backups run in an
 acceptable amount of time, I stay happy.

 - --
  _  _ _  _ ___  _  _  _
 |Y#| |  | |\/| |  \ |\ |  | |Ryan Novosielski - Systems Programmer III
 |$| |__| |  | |__/ | \| _| |[EMAIL PROTECTED] - 973/972.0922 (2-0922)
 \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630
 -BEGIN PGP SIGNATURE-
 Version: GnuPG v1.4.5 (MingW32)
 Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

 iD8DBQFFVLDxmb+gadEcsb4RAhpWAJ0cG8KPP5SmnY8XWyGv80OU8CDCuACeK0zn
 PuKB65lC6TaN7Akq0Cf/XtI=
 =SsjZ
 -END PGP SIGNATURE-


 -
 Using Tomcat but need to do more? Need to support web services, security?
 Get stuff done quickly with pre-integrated technology to make your job
 easier
 Download IBM WebSphere Application Server v.1.0.1 based on Apache 
 Geronimo
 http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
 ___
 Bacula-users mailing list
 Bacula-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/bacula-users

 
 
 What help me get some extra rate was:
 
   Network Buffer Size, but it took me 1 week testing all my clients and
 decide wich value use on each client, before i setup the production server.
 
 Greetings!!!
 
 P.S. My Backup server is FreeBSD to.

Can you offer any guidelines as to how best determine the buffer size?

DAve

-- 
Three years now I've asked Google why they don't have a
logo change for Memorial Day. Why do they choose to do logos
for other non-international holidays, but nothing for
Veterans?

Maybe they forgot who made that choice possible.

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642
___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users


Re: [Bacula-users] Data transfer rates

2006-11-10 Thread pedro moreno
On 11/10/06, DAve [EMAIL PROTECTED] wrote:
pedro moreno wrote: On 11/10/06, Ryan Novosielski [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1
 DAve wrote:  Good morning,  I'm not sure how to interpret the variation in rates. I don't think I  have a problem, though I would expect the backups across the 1gb
 network  to be far faster than they are. The backups across the 1gb network are  all FreeBSD boxes, the backups across the 100mb and the 13mb VLAN are  Windows and RedHat servers. The dir, and the sd clients are running on
  FreeBSD servers dedicated to that task, there are no other processes  running.   Any ideas where I might start looking? There are a couple of things that jump out at me. Compression seems to
 be affecting your rates a certain amount -- that probably makes sense, as it takes some time to do compression. If you can, I'd be curious to see what happens if you turn compression off and try the same numbers.
 That said, your speed issues might be due to A) spooling or B) misconfigured duplex settings on your machine. Check netstat -ian or your platform's equivalent to make sure your
 network settings are OK. As far as the spooling goes, you could try without to see what happens (if you are using it), but basically my take on that effect was that's the way it goes -- you spool and your rate
 drops. I'm not thrilled that I no longer have a good way to check the SD to tape data transfer rate, but as long as the backups run in an acceptable amount of time, I stay happy.
 - --  __ __ __ |Y#| || |\/| |\ |\ || |Ryan Novosielski - Systems Programmer III |$| |__| || |__/ | \| _| |novosirj@
umdnj.edu - 973/972.0922 (2-0922) \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.5 (MingW32) Comment: Using GnuPG with Mozilla - 
http://enigmail.mozdev.org iD8DBQFFVLDxmb+gadEcsb4RAhpWAJ0cG8KPP5SmnY8XWyGv80OU8CDCuACeK0zn PuKB65lC6TaN7Akq0Cf/XtI= =SsjZ -END PGP SIGNATURE-
 - Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job
 easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo 
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642 ___ Bacula-users mailing list 
Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users What help me get some extra rate was:
 Network Buffer Size, but it took me 1 week testing all my clients and decide wich value use on each client, before i setup the production server. Greetings!!! P.S
. My Backup server is FreeBSD to.Can you offer any guidelines as to how best determine the buffer size?DAve--Three years now I've asked Google why they don't have alogo change for Memorial Day. Why do they choose to do logos
for other non-international holidays, but nothing forVeterans?Maybe they forgot who made that choice possible.-Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easierDownload IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642___Bacula-users mailing list
Bacula-users@lists.sourceforge.nethttps://lists.sourceforge.net/lists/listinfo/bacula-users  Edit bacula-sd.conf
, change Network Buffer Size, restart the sd service, them edit bacula-fd.conf for each client restart the fd-service, do 3 backups for each client, get the average. Starting from 8192 to 262144. I didnt found any other way to made this, but this was my test server, the same machine all this was to get the best values and them setup the server to production.
Some NT4 boxes work good with 65536 another 32768, Linux RH9 65536, none of the clients support 262144 even FreeBSD 6.1. Was time-consume, but get me some extra rate.Greetings.
-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnkkid=120709bid=263057dat=121642___
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users