[mira_talk] R: Re: PC with 64 GB RAM: update

  • From: "Davide Sassera (davide.sassera)" <davide.sassera@xxxxxxxx>
  • To: mira_talk@xxxxxxxxxxxxx
  • Date: Mon, 30 Jan 2012 09:37:54 +0100

I also have a 64G machine,

I was able to accurate de novo 5 mil pairs (10 Mil total) solexa reads (100bp). 
When I tried to go 7.5Mil pairs it died saying something about too many skim 
hits if I remember correctly.

I can send the log if needed.

best
D.



Il 30/01/12, Dimitar Kenanov  <dimitar@xxxxxxxxxxxxxx> ha scritto:
> Hi Bastien,
> for now i ran MIRA with a 2M torrent reads and A. Thaliana ref genome with 
> that command:
> 
> mira --project=NON-316c1_17 --job=mapping,est,accurate,iontor -AS:nop=1 
> -SB:bft=gbf:bbq=30 IONTOR_SETTINGS -SB:ads=yes:dsn=athali >& log_assembly.log
> 
> It took 1.15 h and around 26Gb RAM.
> 
> I tried the same data and with that one:
> 
> mira --project=DIV-316c1_16 --job=mapping,genome,accurate,iontor -AS:nop=1 
> -SB:bft=gbf:bbq=30 IONTOR_SETTINGS -SB:ads=yes:dsn=athali
> 
> This one however did not finish correctly. I got the following error message 
> which i have no idea how to interpret:
> ----------- LOG -------------
> "offsets1tos2<0 && abs(offsets1tos2)>=len2 ?"
> 
> ->Thrown: void ADSEstimator::calcNewEstimate(int32 offsets1tos2, uint32 len1, 
> uint32 len2, int32 id1, int32 id2, int8 id1dir, int8 id2dir)
> ->Caught: main
> 
> Aborting process, probably due to an internal error.
> ----------- LOG --------------
> 
> 
> And here is the some info before the error occured:
> 
> 
> ------------- LOG -------------
> 
> Backbone SW checks reverse:
> System memory: 67542007808
> Mem2keepfree: 10131301171
> Used by MIRA: 27041959936
> Mem avail: 30368746701
> rsh increased memtouse to: 30368746701
> tcmalloc: large alloc 2672017408 bytes == 0x57276e000 @
> Can load up to 83500326 skim edges at once.
> Partitioning into 1 blocks.
> We have 79524120 skims in file.
> Localtime: Fri Jan 27 18:25:24 2012
> De-normalising SKIM hits ... (this will take a while)
> Only one block ... first loop ...not loaded yet.
> Loading skims in id range 0 to 2004021
> Localtime: Fri Jan 27 18:25:28 2012
> Loaded 79524120 elements.
> Sorting ... done.
> Localtime: Fri Jan 27 18:25:39 2012
> Writing normalised skimblock 0 (     2.4 GiB) ... done.
> Only one block ... not in first loop, return 0.
> Localtime: Fri Jan 27 18:25:40 2012
> Loading block 0
> Only one block, already loaded.
> Localtime: Fri Jan 27 18:25:41 2012
> Step rail
> Loading block 0
> Only one block, already loaded.
> something is really wrong with that ADSE!
> -922
> 108
> 140
> 1246186
> 594085
> 1
> 1
> adse: id1: 1246186 id2: 594085 le1: 922 re1: 0 le2: 0 re2: 890 ln1: 108 ln2: 
> 140 tl: 1030 ol: 4294966514 dr1: 1 dr2: 1
> ========================== Memory self assessment 
> ==============================
> Running in 64 bit mode.
> 
> Dump from /proc/meminfo
> --------------------------------------------------------------------------------
> MemTotal:       65958992 kB
> MemFree:         9889588 kB
> Buffers:           43640 kB
> Cached:         28225132 kB
> SwapCached:            0 kB
> Active:         39364792 kB
> Inactive:       15763672 kB
> Active(anon):   26729828 kB
> Inactive(anon):   137540 kB
> Active(file):   12634964 kB
> Inactive(file): 15626132 kB
> Unevictable:          36 kB
> Mlocked:              36 kB
> SwapTotal:       1951888 kB
> SwapFree:        1951888 kB
> Dirty:           2485144 kB
> Writeback:             0 kB
> AnonPages:      26859904 kB
> Mapped:           147260 kB
> Shmem:              7648 kB
> Slab:             821940 kB
> SReclaimable:     796700 kB
> SUnreclaim:        25240 kB
> KernelStack:        3392 kB
> PageTables:        62572 kB
> NFS_Unstable:          0 kB
> Bounce:                0 kB
> WritebackTmp:          0 kB
> CommitLimit:    34931384 kB
> Committed_AS:   27618376 kB
> VmallocTotal:   34359738367 kB
> VmallocUsed:      307036 kB
> VmallocChunk:   34359427580 kB
> DirectMap4k:       53248 kB
> DirectMap2M:     3051520 kB
> DirectMap1G:    63963136 kB
> --------------------------------------------------------------------------------
> 
> Dump from /proc/self/status
> --------------------------------------------------------------------------------
> Name:    mira
> State:    R (running)
> Tgid:    4927
> Pid:    4927
> PPid:    3975
> TracerPid:    0
> Uid:    1000    1000    1000    1000
> Gid:    100    100    100    100
> FDSize:    256
> Groups:    11 17 18 19 93 100
> VmPeak:    26408168 kB
> VmSize:    26408164 kB
> VmLck:           0 kB
> VmHWM:    26351184 kB
> VmRSS:    26341524 kB
> VmData:    26377236 kB
> VmStk:         340 kB
> VmExe:        3336 kB
> VmLib:        4580 kB
> VmPTE:       51548 kB
> Threads:    1
> SigQ:    0/515273
> SigPnd:    0000000000000000
> ShdPnd:    0000000000000000
> SigBlk:    0000000000000000
> SigIgn:    0000000000000000
> SigCgt:    0000000180000000
> CapInh:    0000000000000000
> CapPrm:    0000000000000000
> CapEff:    0000000000000000
> CapBnd:    ffffffffffffffff
> Cpus_allowed:    fff
> Cpus_allowed_list:    0-11
> voluntary_ctxt_switches:    63789
> nonvoluntary_ctxt_switches:    33490
> --------------------------------------------------------------------------------
> 
> Information on current assembly object:
> 
> AS_readpool: 2004022 reads.
> AS_contigs: 0 contigs.
> AS_bbcontigs: 5 contigs.
> Mem used for reads: 6526343832 (6.1 GiB)
> 
> Memory used in assembly structures:
>                                            Eff. Size   Free cap. LostByAlign
>      AS_writtenskimhitsperid:    2004022       8 MiB         0 B         0 B
>                AS_skim_edges:   79524120     2.5 GiB     121 MiB         0 B
>                  AS_adsfacts:          0        24 B         0 B         0 B
>           AS_confirmed_edges:          0        24 B         0 B         0 B
>    AS_permanent_overlap_bans:          1        24 B         0 B         0 B
>               AS_readhitmiss:          0        24 B         0 B         0 B
>             AS_readhmcovered:          0        24 B         0 B         0 B
>                 AS_count_rhm:          0        24 B         0 B         0 B
>                  AS_clipleft:    2004022       8 MiB         0 B         0 B
>                 AS_clipright:    2004022       8 MiB         0 B         0 B
>                  AS_used_ids:    2004022       2 MiB         0 B         2 B
>               AS_multicopies:    2004022       2 MiB         0 B         2 B
>             AS_hasmcoverlaps:          0       2 MiB       2 MiB         2 B
>        AS_maxcoveragereached:    2004022       8 MiB         0 B         0 B
>        AS_coverageperseqtype:          0        24 B         0 B         0 B
>            AS_istroublemaker:    2004022       2 MiB         0 B         2 B
>                  AS_isdebris:    2004022       2 MiB         0 B         2 B
>           AS_needalloverlaps:    2004022       2 MiB        10 B         0 B
>     AS_readsforrepeatresolve:          0        40 B         0 B         0 B
>                 AS_allrmbsok:          0       8 MiB       8 MiB         0 B
>         AS_probablermbsnotok:          0       8 MiB       8 MiB         0 B
>             AS_weakrmbsnotok:          0       8 MiB       8 MiB         0 B
>           AS_readmaytakeskim:    2004022       2 MiB        10 B         0 B
>                AS_skimstaken:   79524120      76 MiB        40 B         0 B
>           AS_numskimoverlaps:    2004022       8 MiB         0 B         0 B
>        AS_numleftextendskims:    2004022       8 MiB         0 B         0 B
>          AS_rightextendskims:    2004022       8 MiB         0 B         0 B
>       AS_skimleftextendratio:    2004022       2 MiB         0 B         2 B
>      AS_skimrightextendratio:    2004022       2 MiB         0 B         2 B
>              AS_usedtmpfiles:         10       336 B         0 B         0 B
> Total: 9376076632 (8.7 GiB)
> 
> ================================================================================
> Dynamic allocs: 6
> Align allocs: 22
> ----------------------- LOG ------------------------
> 
> 
> Any idea what might have gone wrong?
> 
> Dimitar
> 
> 
> On 01/28/2012 11:30 PM, Bastien Chevreux wrote:
> >On Jan 27, 2012, at 6:30 , Dimitar Kenanov wrote:
> >   
> >>i got the new PC. It has CPU - i7 3930K and 64GB RAM. MBoard - Gigabyte 
> >>X79-UD5.
> >>Well it works. From my Slackware linux i can see all of the RAM. I attach a 
> >>JPG where in a window which runs 'htop' program one can see all the CPU 
> >>threads and the RAM amount.
> >>Unfortunately didnt test it with MIRA yet. But i will for sure and report 
> >>how the machine behaves.
> >>     
> >Uh, that thing is making me jealous. Really :-)
> >
> >Have you tried whether a single process can allocate and use>32 GiB?
> >
> >B.
> >
> >
> >   
> 
> 
> 
> -- 
> You have received this mail because you are subscribed to the mira_talk 
> mailing list. For information on how to subscribe or unsubscribe, please 
> visit http://www.chevreux.org/mira_mailinglists.html
> 
>

Other related posts:

  • » [mira_talk] R: Re: PC with 64 GB RAM: update - Davide Sassera (davide.sassera)