Peter Eriksson wrote:
On Wed, 24 May 2006, Dan Trainor wrote:
I couldn't help but notice you mentioning that it had taken over two
whole days to recover 360G of data. Are you kidding?
Now the 2 days have passed and this time I was able to not make
any mistakes and have started the actual recover process...
Current estimates indicates that it processes about 1GB per hour...
So it'll take about 11 days to recover the 305GB / 5 million files.
Even if the incore tree building is very slow due to bad/missing
indexes in the MySQL database I think this part of the recover shouldn't
really be effected by that problem?
I wonder what the speed limiting factor is this time...
Server: Sun Ultra 60 with dual 360MHz processors, 2GB RAM
Tape drives: Quantum DLT-7000 tape drives
Robot: Sun StorEdge L-11000 (ATL P3000)
The restore is running locally on the backup server and writes to a
NFS-mounted file system from another server via Gigabit Ethernet
so that shouldn't be the limiting factor...
"top" output:
last pid: 17097; load averages: 0.16, 0.16, 0.16
16:06:05
42 processes: 41 sleeping, 1 on cpu
CPU states: 94.6% idle, 1.4% user, 4.0% kernel, 0.0% iowait, 0.0% swap
Memory: 2048M real, 34M free, 1737M swap in use, 1588M swap free
PID USERNAME THR PRI NICE SIZE RES STATE TIME CPU COMMAND
11507 root 3 59 0 5800K 2456K sleep 18:09 2.19% bacula-fd
11613 daemon 18 59 0 4368K 2256K sleep 1:26 1.37% nfsmapid
17097 root 1 59 0 2168K 1680K cpu/2 0:00 1.14% top
11509 root 15 59 0 1385M 1050M sleep 110.3H 0.31% bacula-dir
11501 root 3 59 0 6640K 2696K sleep 224:59 0.20% bacula-sd
572 mysql 11 59 0 309M 289M sleep 250:49 0.11% mysqld
16833 root 1 59 0 4112K 2088K sleep 0:18 0.05% sshd
16921 root 2 59 0 5424K 2352K sleep 0:11 0.03% bconsole
Ah well. I'll just let it trundle along I think and we'll see
how long it takes.
Hmm.. Perhaps the tape drive is doing a lot of start-stop operations?
(Can't really check it right now - I'm 30km away from it right now :-)
Top shows over half of your 3 GB swap space being utilized. Almost as
much memory is in swap as is in physical RAM. Swap is perhaps being
thrashed.
Josh Fisher
-------------------------------------------------------
All the advantages of Linux Managed Hosting--Without the Cost and Risk!
Fully trained technicians. The highest number of Red Hat certifications in
the hosting industry. Fanatical Support. Click to learn more
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
