Skipping queue run -- load average too high
Ugo Bellavance
ugob at CAMO-ROUTE.COM
Wed Dec 10 14:33:37 GMT 2003
> -----Message d'origine-----
> De : Harry Hanson [mailto:harryh at CET.COM]
> Envoyé : Wednesday, December 10, 2003 12:47 AM
> À : MAILSCANNER at JISCMAIL.AC.UK
> Objet : Re: Skipping queue run -- load average too high
>
>
> > > -----Message d'origine-----
> > > De : Harry Hanson [mailto:harryh at CET.COM] Envoyé : Tuesday,
> > December
> > > 09, 2003 9:17 PM À : MAILSCANNER at JISCMAIL.AC.UK Objet : Re:
> > Skipping
> > > queue run -- load average too high
> > >
> > >
> > > Dual 1.6 xeon, 1gb memory, 10k rpm u320 scsi drives in
> > raid0 config,
> > > 1gb ram.
> >
> > What is your daily mail volume?
>
> Here's and example from midnight to current time:
>
> Log starts at Dec 9 00 00:01 and ends at Dec 9 21 41:42
>
> Total bytes transferred: 1496649
> Total bytes In: 0
> Messages Out: 33742
> Messages In: 0
>
> Messages per hour (each dot is 181 messages)
> ________________
> 0: 827 ....
> 1: 838 ....
> 2: 745 ....
> 3: 793 ....
> 4: 778 ....
> 5: 804 ....
> 6: 805 ....
> 7: 856 ....
> 8: 781 ....
> 9: 839 ....
> 10: 748 ....
> 11: 871 ....
> 12: 799 ....
> 13: 778 ....
> 14: 823 ....
> 15: 9056 ..................................................
> 16: 6625 ....................................
> 17: 1462 ........
> 18: 1478 ........
> 19: 1383 .......
> 20: 1364 .......
> 21: 1056 .....
> 22: 0
> 23: 0
>
>
> > >
> > > Freebsd 5.1
> > >
> > > It would seem to me either an I/O issue, as the cpu rarely
> > exceeds 5%
> > > useage, or perhaps I have something set too low?
> > >
> > If you've got vmstat on freebsd, use it... type vmstat 1.
> >
> > this will tell you if you have processes waiting for I/O.
> > (see man vmstat)
>
> My apoligies; I am rather a newbie with freebsd; this is my
> first. Here's a
> current snapshot (tho I have to read the man page more; it scrolls
> coninually and not sure how to interpret the output):
You have to stop it with ctrl-c. To interpret the output, see the manpage for vmstat. The first column I check is usually the second one (b). And in your case, is seems ok, because it is 0 or 1 (there is one or zero process waiting for I/O). I don't have a freebsd box here, so I can't interpret everything.
>
>
> procs memory page disks
> faults cpu
> r b w avm fre flt re pi po fr sr aa0 aa1 in
> sy cs us sy
> id
> 0 0 0 224380 314004 11724 0 0 0 10377 0 17 0
> 533 0 2964 34
> 28 38
> 0 1 0 224896 313344 147 0 0 0 41 0 3 57 688
> 0 1570 2 11
> 87
> 0 0 1 221424 315056 48 0 0 0 469 0 0 5 445
> 0 595 0 4
> 96
> 0 1 1 224296 313860 331 0 0 0 80 0 0 26 553
> 0 927 1 5
> 94
> 0 0 1 224332 313400 124 0 0 0 43 0 0 32 626
> 0 1106 1 5
> 94
> 0 0 1 226080 312884 5447 0 0 0 4632 0 8 32 660
> 0 1577 16 13
> 71
> 0 0 1 222444 314720 178 0 0 0 685 0 0 56 714
> 0 1401 1 8
> 91
> 0 0 1 222940 314612 292 0 0 0 305 0 0 30 492
> 0 782 0 4
> 96
> 0 0 1 223712 314196 113 0 0 0 13 0 0 0 404
> 0 488 0 4
> 96
> 0 0 3 233728 311680 5302 0 0 0 4153 0 10 0 490
> 0 981 9 15
> 76
> 0 0 1 220824 315280 3880 0 0 0 4300 0 2 0 593
> 0 1151 14 10
> 76
> 0 0 1 219360 315836 266 0 0 0 436 0 0 24 553
> 0 929 1 4
> 95
> 0 2 1 220908 315104 468 0 0 0 315 0 0 7 607
> 0 1027 1 7
> 92
> 0 0 1 219352 316120 256 0 0 0 557 0 0 43 602
> 0 1086 1 5
> 94
> 0 1 2 224964 313308 6070 0 0 0 4839 0 20 9 518
> 0 1083 11 16
> 72
> 0 0 0 221508 315068 3038 0 0 0 3045 0 0 17 641
> 0 1304 16 8
> 76
> 0 0 0 221552 315000 135 0 0 0 130 0 0 11 496
> 0 744 0 6
> 94
> 0 0 0 220272 315616 117 0 0 0 303 0 11 36 540
> 0 914 1 4
> 95
> 0 0 0 220264 315652 362 0 0 0 387 0 0 9 517
> 0 777 1 4
> 95
> 0 1 1 223036 314032 4288 0 0 0 3483 0 12 19 567
> 0 1294 11 13
> 76
> 0 0 0 218800 316340 2958 0 0 0 3155 0 0 57 797
> 0 3355 13 18
> 69
> 0 0 0 215380 318212 488 0 0 0 971 0 0 9 540
> 0 855 0 7
> 92
> procs memory page disks
> faults cpu
>
More information about the MailScanner
mailing list