|
Re: Troubleshooting a slowdown problem?
2004-06-17 19:05:52
--On Thursday, June 17, 2004 17:08:46 -0400 KEVIN ZEMBOWER <KZEMBOWE AT jhuccp
DOT org> wrote:
> Ohh, it ended just as I pressed 'send.' Here's the daily report:
> These dumps were to tape DailySet104.
> The next tape Amanda expects to use is: DailySet105.
>
> FAILURE AND STRANGE DUMP SUMMARY:
> admin //db/f$/inetsrv/webpub/images lev 1 STRANGE
> admin //db/e$ lev 1 STRANGE
> admin //db/f$ lev 1 STRANGE
> admin //db/c$ lev 0 STRANGE
>
>
> STATISTICS:
> Total Full Daily
> -------- -------- --------
> Estimate Time (hrs:min) 0:04
> Run Time (hrs:min) 21:03
> Dump Time (hrs:min) 29:18 28:55 0:23
> Output Size (meg) 10446.5 7946.7 2499.8
> Original Size (meg) 15747.4 13247.4 2500.0
> Avg Compressed Size (%) 58.5 58.5 23.5 (level:#disks ...)
> Filesystems Dumped 17 13 4 (1:4)
> Avg Dump Rate (k/s) 101.4 78.2 1882.1
>
> Tape Time (hrs:min) 2:52 2:27 0:25
> Tape Size (meg) 10446.7 7946.9 2499.8
> Tape Used (%) 90.3 68.7 21.6 (level:#disks ...)
> Filesystems Taped 17 13 4 (1:4)
> Avg Tp Write Rate (k/s) 1038.9 922.9 1730.4
>
> USAGE BY TAPE:
> Label Time Size % Nb
> DailySet104 2:52 10446.7 90.3 17
>
>
> FAILED AND STRANGE DUMP DETAILS:
>
> /-- admin //db/f$/inetsrv/webpub/images lev 1 STRANGE
> sendbackup: start [admin://db/f$/inetsrv/webpub/images level 1]
> sendbackup: info BACKUP=/usr/bin/smbclient
> sendbackup: info RECOVER_CMD=/usr/bin/smbclient -f... -
> sendbackup: info end
> ? INFO: Debug class all level = 1 (pid 4428 from pid 4428)
>| added interface ip=172.16.2.7 bcast=172.16.255.255 nmask=255.255.0.0
>| tar: dumped 5 files and directories
>| Total bytes written: 22528
> sendbackup: size 22
> sendbackup: end
> \--------
>
> /-- admin //db/e$ lev 1 STRANGE
> sendbackup: start [admin://db/e$ level 1]
> sendbackup: info BACKUP=/usr/bin/smbclient
> sendbackup: info RECOVER_CMD=/usr/bin/smbclient -f... -
> sendbackup: info end
> ? INFO: Debug class all level = 1 (pid 4462 from pid 4462)
>| added interface ip=172.16.2.7 bcast=172.16.255.255 nmask=255.255.0.0
>| tar: dumped 9 files and directories
>| Total bytes written: 1536
> sendbackup: size 2
> sendbackup: end
> \--------
>
> /-- admin //db/f$ lev 1 STRANGE
> sendbackup: start [admin://db/f$ level 1]
> sendbackup: info BACKUP=/usr/bin/smbclient
> sendbackup: info RECOVER_CMD=/usr/bin/smbclient -f... -
> sendbackup: info end
> ? INFO: Debug class all level = 1 (pid 4482 from pid 4482)
>| added interface ip=172.16.2.7 bcast=172.16.255.255 nmask=255.255.0.0
>| tar: dumped 215 files and directories
>| Total bytes written: 2620979200
> sendbackup: size 2559550
> sendbackup: end
> \--------
>
> /-- admin //db/c$ lev 0 STRANGE
> sendbackup: start [admin://db/c$ level 0]
> sendbackup: info BACKUP=/usr/bin/smbclient
> sendbackup: info RECOVER_CMD=/usr/bin/smbclient -f... -
> sendbackup: info end
> ? INFO: Debug class all level = 1 (pid 4486 from pid 4486)
>| added interface ip=172.16.2.7 bcast=172.16.255.255 nmask=255.255.0.0
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\system
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\software
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\default
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file
> \WINNT\system32\config\software.LOG (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\default.LOG
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\SECURITY
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file
> \WINNT\system32\config\SECURITY.LOG (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\SYSTEM.ALT
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\SAM
> (\WINNT\system32\config\)
> ? ERRDOS - ERRbadshare opening remote file \WINNT\system32\config\SAM.LOG
> (\WINNT\system32\config\)
> ? ERRDOS - ERRnoaccess opening remote file \pagefile.sys (\)
>| tar: dumped 5941 files and directories
>| Total bytes written: 470870016
> sendbackup: size 459834
> sendbackup: end
> \--------
>
>
> NOTES:
> planner: Last full dump of admin://db/f$ on tape DailySet105 overwritten in
> 1 run.
This is a big problem. After your next run you will no longer be able to do a
full restore of this filesystem. This is caused by the line below.
> planner: admin //db/f$ 20040616 0 [dump larger than tape, 12838742 KB, full
> dump delayed]
Amanda knows its time to do a full, but is unable to because the dump
won't fit on a single tape. You either need to split this disklist entry
into 2 (or more) subdirectories or experiment with the tape spanning code
still in development (or get bigger tapes).
> planner: Full dump of centernet:sda2 promoted from 2 days ahead.
> planner: Full dump of mailinglists:hda2 promoted from 1 day ahead.
> planner: Full dump of admin:sdb1 promoted from 2 days ahead.
> planner: Full dump of mailinglists:hda7 promoted from 1 day ahead.
> planner: Full dump of centernet:sda9 promoted from 2 days ahead.
> planner: Full dump of admin://db/c$ promoted from 1 day ahead.
> planner: Full dump of mailinglists:hda1 promoted from 2 days ahead.
> planner: Full dump of centernet:sda3 promoted from 2 days ahead.
> planner: Full dump of admin:sda1 promoted from 2 days ahead.
> planner: Full dump of centernet:sda7 promoted from 2 days ahead.
> planner: Full dump of centernet:sda1 promoted from 2 days ahead.
> taper: tape DailySet104 kb 10697984 fm 17 [OK]
>
>
> DUMP SUMMARY:
> DUMPER STATS TAPER STATS
> HOSTNAME DISK L ORIG-KB OUT-KB COMP% MMM:SS KB/s MMM:SS KB/s
> -------------------------- --------------------------------- ------------
> admin //db/c$ 0 462720 462720 -- 4:191788.1 5:141475.5
> admin //db/e$ 1 10 32 -- 0:09 1.1 0:002839.9
> admin //db/f$ 1 25596602559680 -- 9:544312.1 24:391730.4
> admin -pub/images 1 30 32 -- 12:34 0.0 0:002822.6
> admin sda1 0 3410 3424 -- 0:013081.9 0:031014.4
> admin sda3 0 54646573821088 69.9 70:36 902.1 70:37 901.9
> admin sdb1 0 24270 24288 -- 0:055047.4 0:38 640.4
All the rates on admin look good.
> centernet sda1 0 5670 4864 85.8 1:54 42.5 0:05 887.7
> centernet sda2 0 1482550 715744 48.3 306:39 38.9 13:14 901.6
> centernet sda3 0 317710 110912 34.9 37:28 49.3 2:00 920.9
> centernet sda5 0 20678301818176 87.9 796:47 38.0 33:21 908.7
> centernet sda6 1 310 96 31.0 0:04 20.2 0:003080.0
> centernet sda7 0 6090 576 9.5 0:15 37.9 0:003035.6
> centernet sda9 0 580430 30400 5.2 12:29 40.6 0:33 910.4
> mailinglists hda1 0 2861 2208 77.2 0:54 40.8 0:02 984.9
> mailinglists hda2 0 983941 399360 40.6 171:46 38.7 7:27 892.9
> mailinglists hda7 0 2163157 743840 34.4 332:00 37.3 13:42 905.2
The dumper rates on these others are pretty poor. Do you have a duplex
mis-match on your server, client or switch? If you ftp a fair-sized file
both ways between admin and centernet, what transfer rates do you get?
I get a little over twice that rate on the clients I backup over a 4.5 megabit
VPN WAN link, and even my slowest local client (an ancient MIPS-based box) can
dump at around 100KB/s. Most of my clients dump (client compress tar, actually)
in the 1000-3000KB/s range.
You might want to set columnspec in your config so the numbers don't all
run together.
Frank
>
> (brought to you by Amanda version 2.4.4)
>
> I'll take a closer look at this tomorrow. You all have a good evening.
>
> -Kevin
--
Frank Smith fsmith AT hoovers DOT com
Sr. Systems Administrator Voice: 512-374-4673
Hoover's Online Fax: 512-374-4501
|
|
|
