Hi,
I'm configuring Amanda version 2.5.0-2-4 on Red Hat Enterprise Linux 5
Here's the OS kernel revision:
-bash-3.1$ uname -a
Linux lto3.gc.cuny.edu 2.6.18-8.1.4.el5 #1 SMP Fri May 4 22:15:13 EDT
2007 i686 i686 i386 GNU/Linux
Earlier in the day I ran amtapetype (more specifically I ran
nohup /usr/sbin/amtapetype -o -e 400g -f /dev/nst1 -t LTO3-400G >
mytype 2>&1 &
I'm using an Exabyte Magnum 224 tape library with an LTO-3 drive; to configure
a tapetype for amanda, I ran amtapetype. The tapes are Ultrium LTO-3
400GB uncompressed, 800GB compressed.
Of course, before the exercise I ran
mt -f /dev/nst1 compression off
(the use of off isn't documented in the man pages, so this is worth something in
consulting fees--I accept PayPal).
Anyway, amtapetype gave me an answer, but not before complaining:
-bash-3.1$ more mytype
Writing 2048 Mbyte compresseable data: 30 sec
Writing 2048 Mbyte uncompresseable data: 30 sec
Estimated time to write 2 * 409600 Mbyte: 12000 sec = 3 h 20 min
wrote 12845056 32Kb blocks in 98 files in 5659 seconds (short write)
wrote 12910592 32Kb blocks in 197 files in 6025 seconds (short write)
*** glibc detected *** /usr/sbin/amtapetype: free(): invalid pointer: 0xb7e5e000
***
======= Backtrace: =========
/lib/libc.so.6[0x53df7d]
/lib/libc.so.6(cfree+0x90)[0x5415d0]
/usr/sbin/amtapetype(main+0xa04)[0xa20e54]
/lib/libc.so.6(__libc_start_main+0xdc)[0x4eddec]
/usr/sbin/amtapetype[0xa1fc11]
======= Memory map: ========
00110000-0011a000 r-xp 00000000 08:02 203084 /usr/lib/libamtape-2.5.0p2.so
0011a000-0011b000 rwxp 00009000 08:02 203084 /usr/lib/libamtape-2.5.0p2.so
0011b000-0013c000 r-xp 00000000 08:02 220202 /usr/lib/libamanda-2.5.0p2.so
0013c000-0013d000 rwxp 00021000 08:02 220202 /usr/lib/libamanda-2.5.0p2.so
0013d000-0015c000 rwxp 0013d000 00:00 0
0015c000-00167000 r-xp 00000000 08:02 2615560 /lib/libgcc_s-4.1.1-20070105.so
.1
00167000-00168000 rwxp 0000a000 08:02 2615560 /lib/libgcc_s-4.1.1-20070105.so
.1
00295000-002ae000 r-xp 00000000 08:02 2615538 /lib/ld-2.5.so
002ae000-002af000 r-xp 00018000 08:02 2615538 /lib/ld-2.5.so
002af000-002b0000 rwxp 00019000 08:02 2615538 /lib/ld-2.5.so
003cd000-003f2000 r-xp 00000000 08:02 2615541 /lib/libm-2.5.so
003f2000-003f3000 r-xp 00024000 08:02 2615541 /lib/libm-2.5.so
003f3000-003f4000 rwxp 00025000 08:02 2615541 /lib/libm-2.5.so
004c5000-004d4000 r-xp 00000000 08:02 2615552 /lib/libresolv-2.5.so
004d4000-004d5000 r-xp 0000e000 08:02 2615552 /lib/libresolv-2.5.so
004d5000-004d6000 rwxp 0000f000 08:02 2615552 /lib/libresolv-2.5.so
004d6000-004d8000 rwxp 004d6000 00:00 0
004d8000-0060f000 r-xp 00000000 08:02 2615539 /lib/libc-2.5.so
0060f000-00611000 r-xp 00137000 08:02 2615539 /lib/libc-2.5.so
00611000-00612000 rwxp 00139000 08:02 2615539 /lib/libc-2.5.so
00612000-00615000 rwxp 00612000 00:00 0
00847000-00848000 r-xp 00847000 00:00 0 [vdso]
00a1f000-00a22000 r-xp 00000000 08:02 209083 /usr/sbin/amtapetype
00a22000-00a23000 rwxp 00002000 08:02 209083 /usr/sbin/amtapetype
00ad5000-00ad8000 r-xp 00000000 08:02 2615750 /lib/libtermcap.so.2.0.8
00ad8000-00ad9000 rwxp 00002000 08:02 2615750 /lib/libtermcap.so.2.0.8
00f2d000-00f3f000 r-xp 00000000 08:02 2615551 /lib/libnsl-2.5.so
00f3f000-00f40000 r-xp 00012000 08:02 2615551 /lib/libnsl-2.5.so
00f40000-00f41000 rwxp 00013000 08:02 2615551 /lib/libnsl-2.5.so
00f41000-00f43000 rwxp 00f41000 00:00 0
08b8b000-08bac000 rw-p 08b8b000 00:00 0
b7d00000-b7d21000 rw-p b7d00000 00:00 0
b7d21000-b7e00000 ---p b7d21000 00:00 0
b7e5d000-b7f62000 rw-p b7e5d000 00:00 0
b7f75000-b7f76000 rw-p b7f75000 00:00 0
bff03000-bff18000 rw-p bff03000 00:00 0 [stack]
define tapetype LTO3-400G {
comment "just produced by tapetype prog (hardware compression off)"
length 402432 mbytes
filemark 0 kbytes
speed 70602 kps
}
Now this looks tapetype entry looks decent, but something tells me
I need to modify my setup, or other amanda calls might suffer the same fate.
Does this look peculiar to my system? Is there a buglet in amtapetype?
Should I download the source and compile it instead of using yum?
FL
PS. I have to say that amanda has been the most robust backup system I've used.
It saved the day a couple of times in the past month, most recently
when a routine
up2date on my red hat system hung up. The only recourse: reboot the machine.
The consequence: a corrupted LDAP database! No more user logins! But amanda
had 11 good backups. In another case, the level 0 failed because it grew too big
(subsequently addressed) but an incrementals was enough.
|
