Hello,
Yes, you have a sharp eye :-). Once in the past the machine had 1GB RAM
and its swap was sized accordingly to 2GB. Last week I let additional 2
GB RAM put in, the swap remained the same, after some OOM I added an
additional 6 GB swap space. But in theory, and this is my question, 1GB
ought to be enough, am I right? For the record:
Mem: 3092292k total, 3070416k used, 21876k free, 4992k buffers
Swap: 6197568k total, 1329312k used, 4868256k free, 664764k cached
4498 root 16 0 3591m 1.9g 1484 S 0.0 64.5 111:15.21 bacula-dir
20515 mysql 16 0 292m 266m 3024 S 0.3 8.8 103:09.24 mysqld
28746 root 16 0 111m 11m 1348 S 1.7 0.4 127:13.07 bacula-sd
So it is growing and growing.
--tef
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [20070318 10:18]:
> Hello,
>
> I'm definitely not an expert on OS configuration, but from what I see below,
> I
> am either misinterpreting the output, or your OS has a configuration problem.
> >From what I see from the output of "free", you have 3GB of memory and 6GB of
> swap space. That is good.
>
> However, when the kernel kills bacula-dir, the output says:
>
> Free swap = 0kB
> Total swap = 2003272kB
>
> which I take to mean that it is using 2GB of swap space not the 6 that are
> shown above. Obviously 2GB of swap space is totally inadequate if you have
> 3GB of memory. I recommend checking what swap you have allocated on which
> devices, and if you have several swap devices, make sure that they are all
> turned on.
>
> Regards,
>
> Kern
>
> On Sunday 18 March 2007 08:57, SITKEI Attila wrote:
> > Dear list,
> >
> > I use bacula since one year and found to be an excellent backup
> > solution, thank you for it. We currently backup about 120 client using
> > it and plan to migrate to other backups as well.
> >
> > The storage backends are two storages (so no tape are used) both from
> > Transtec, in 4 and 8 TB size, sharing a common lvm, formatted using xfs
> > (not an ideal solution, I know, but had legacy tasks). The devices are
> > working properly at a very low usage (iostat shows up to 15%). In the
> > past, MySQL db (MyISAM backend, inno is completely disabled) was located
> > on a filesystem formatted on the built-in CCISS hardware array, but the
> > IO was slow in some cases so I moved it to the storages. The used
> > SCSI-controller:
> >
> > 0000:06:01.0 SCSI storage controller: LSI Logic / Symbios Logic 53c1030
> > PCI-X Fusion-MPT Dual Ultra320 SCSI (rev 07)
> > NENO(~)$ cat /etc/SuSE-release
> > SUSE LINUX Enterprise Server 9 (x86_64)
> > VERSION = 9
> > PATCHLEVEL = 3
> >
> > NENO(~)$ cat /proc/cpuinfo
> > processor : 0
> > vendor_id : GenuineIntel
> > cpu family : 15
> > model : 4
> > model name : Intel(R) Xeon(TM) CPU 3.20GHz
> > stepping : 1
> > cpu MHz : 3200.243
> > cache size : 1024 KB
> > [...]
> >
> > only one CPU. The machine is a HP DL380 G4 anyway.
> >
> >
> > I had some minor problems using the software, these originated from
> > my ignorance I mean, but for now this problem seems to be something
> > different. Bacula-web says:
> >
> > Total clients: 119 Total bytes stored: 4.53 TB
> > Total files: 5591153 Database size: 3.61 GB
> >
> > The only problem is that the machine that runs bacula-dir keeps having
> > OOM events (out of memory); currently I see these lines is top:
> >
> > 4498 root 16 0 1963m 1.5g 1572 S 0.0 49.5 97:10.62 bacula-dir
> > 20515 mysql 16 0 292m 277m 3688 S 0.0 9.2 93:45.44 mysqld
> > 28746 root 16 0 128m 16m 1356 S 1.0 0.5 112:26.29 bacula-sd
> >
> > while the normal memory usage of -dir is around 300MB, -sd used to eat
> > 100. MySQL is constantly using 300. The machine:
> >
> > Linux neno 2.6.16 #2 SMP Wed Mar 29 08:13:36 CEST 2006 x86_64 x86_64
> > x86_64 GNU/Linux
> >
> > NENO(~)$ free
> > total used free shared buffers
> > cached
> > Mem: 3092292 3070944 21348 0 5940
> > 990736
> > -/+ buffers/cache: 2074268 1018024
> > Swap: 6197568 16 6197552
> >
> > NENO(~)$ rpm -qa | grep bacula
> > bacula-mysql-2.0.3-1
> > bacula-mtx-2.0.3-1
> > bacula-updatedb-2.0.3-1
> >
> > (I used bacula 1.38.10 until the last week but upgraded to give the new
> > version a try)
> >
> > Here a dmesg snippet:
> >
> > Call Trace: <ffffffff801496de>{out_of_memory+60}
> > <ffffffff8014b766>{__alloc_pages+524}
> > <ffffffff80147ca3>{generic_file_buffered_write+440}
> > <ffffffff803111e0>{tcp_rcv_established+1821}
> > <ffffffff8012d9ce>{current_fs_time+77}
> > <ffffffff80183175>{__mark_inode_dirty+230}
> > <ffffffff802404f1>{__up_write+20}
> > <ffffffff80227bf0>{xfs_write+1695}
> > <ffffffff802e2708>{do_sock_read+154}
> > <ffffffff802e3bad>{sock_aio_read+79}
> > <ffffffff801471bc>{filemap_nopage+379}
> > <ffffffff802242fd>{linvfs_aio_write+103}
> > <ffffffff80163c0a>{do_sync_write+201}
> > <ffffffff8013bafd>{autoremove_wake_function+0}
> > <ffffffff801645bf>{vfs_write+175}
> > <ffffffff8016471b>{sys_write+69}
> > <ffffffff8010a762>{system_call+126}
> > Mem-info:
> > DMA per-cpu:
> > cpu 0 hot: high 0, batch 1 used:0
> > cpu 0 cold: high 0, batch 1 used:0
> > DMA32 per-cpu:
> > cpu 0 hot: high 186, batch 31 used:34
> > cpu 0 cold: high 62, batch 15 used:61
> > Normal per-cpu: empty
> > HighMem per-cpu: empty
> > Free pages: 19304kB (0kB HighMem)
> > Active:377849 inactive:378436 dirty:0 writeback:0 unstable:0 free:4826
> > slab:4011 mapped:755464 pagetables:
> > 3137
> > DMA free:12080kB min:24kB low:28kB high:36kB active:136kB inactive:124kB
> > present:12008kB pages_scanned:334
> > all_unreclaimable? yes
> > lowmem_reserve[]: 0 3014 3014 3014
> > DMA32 free:7224kB min:7012kB low:8764kB high:10516kB active:1511260kB
> > inactive:1513620kB present:3086512kB
> > pages_scanned:2750569 all_unreclaimable? no
> > lowmem_reserve[]: 0 0 0 0
> > Normal free:0kB min:0kB low:0kB high:0kB active:0kB inactive:0kB
> > present:0kB pages_scanned:0 all_unreclaim
> > able? no
> > lowmem_reserve[]: 0 0 0 0
> > HighMem free:0kB min:128kB low:128kB high:128kB active:0kB inactive:0kB
> > present:0kB pages_scanned:0 all_un
> > reclaimable? no
> > lowmem_reserve[]: 0 0 0 0
> > DMA: 0*4kB 0*8kB 1*16kB 1*32kB 2*64kB 1*128kB 2*256kB 0*512kB 1*1024kB
> > 1*2048kB 2*4096kB = 12080kB
> > DMA32: 58*4kB 52*8kB 5*16kB 1*32kB 9*64kB 2*128kB 0*256kB 1*512kB
> > 1*1024kB 0*2048kB 1*4096kB = 7224kB
> > Normal: empty
> > HighMem: empty
> > Swap cache: add 856241, delete 855934, find 1489150/1525433, race 0+0
> > Free swap = 0kB
> > Total swap = 2003272kB
> > Free swap: 0kB
> > 786419 pages of RAM
> > 13346 reserved pages
> > 6966 pages shared
> > 307 pages swap cached
> > Out of Memory: Kill process 20474 (bacula-dir) score 74606 and children.
> > Out of memory: Killed process 20474 (bacula-dir).
> > oom-killer: gfp_mask=0x201d2, order=0
> >
> >
> > I can live with the situation, only linux keeps killing applications at
> > arbitrary times. Any suggestions?
> >
> >
> >
> > TIA
> >
> >
> >
> > --tef
> >
> >
> > -------------------------------------------------------------------------
> > 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.php&p=sourceforge&CID=DEVDEV
> > _______________________________________________
> > Bacula-users mailing list
> > Bacula-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/bacula-users
> >
-------------------------------------------------------------------------
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.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
