Re: 2.4.9-ac15 painfully sluggish

From: Pau Aliagas (linux4u@wanadoo.es)
Date: Tue Sep 25 2001 - 13:01:51 EST


On Tue, 25 Sep 2001, Rik van Riel wrote:

> > [pau@pau pau]$ vmstat -n 5
> > procs memory swap io system cpu
> > r b w swpd free buff cache si so bi bo in cs us sy id
> > 2 6 1 2392 3040 476 9644 5 7 585 17 170 479 85 6 9
> > 2 12 0 2520 2800 408 9028 25 27 1425 35 208 184 92 6 2
> > 2 4 0 2596 2812 408 9500 15 22 1174 24 204 304 79 6 16
>
> Wow, these are a LOT of reads and a very small cache.

> Here things get "interesting" ...
>
> total memory: 127 MB
> The memory taken by all your processes: 55 MB
> buffer + cache: 10 MB
> -----------------------------------------------------
> missing: 62 MB
>
> It would be interesting to get a listing of /proc/slabinfo,
> in particular those lines which have a large number in the
> 4th or 5th column...

1. Before starting evolution:
slabinfo - version: 1.1
kmem_cache 59 68 112 2 2 1
ip_conntrack 11 33 352 3 3 1
ip_fib_hash 12 113 32 1 1 1
uhci_urb_priv 1 67 56 1 1 1
ip_mrt_cache 0 0 96 0 0 1
tcp_tw_bucket 1 40 96 1 1 1
tcp_bind_bucket 12 113 32 1 1 1
tcp_open_request 0 0 64 0 0 1
inet_peer_cache 1 59 64 1 1 1
ip_dst_cache 41 100 192 5 5 1
arp_cache 2 30 128 1 1 1
blkdev_requests 512 520 96 13 13 1
dnotify cache 0 0 20 0 0 1
file lock cache 4 42 92 1 1 1
fasync cache 1 202 16 1 1 1
uid_cache 7 113 32 1 1 1
skbuff_head_cache 129 240 160 10 10 1
sock 300 333 832 37 37 2
sigqueue 1024 1044 132 36 36 1
kiobuf 0 0 8768 0 0 4
cdev_cache 23 118 64 2 2 1
bdev_cache 6 59 64 1 1 1
mnt_cache 16 59 64 1 1 1
inode_cache 1036 2178 416 242 242 1
dentry_cache 1019 2730 128 91 91 1
dquot 0 0 128 0 0 1
filp 3116 3120 96 78 78 1
names_cache 0 1 4096 0 1 1
buffer_head 3816 5240 96 131 131 1
mm_struct 83 96 160 4 4 1
vm_area_struct 4536 4897 64 83 83 1
fs_cache 82 118 64 2 2 1
files_cache 82 90 416 10 10 1
signal_act 87 90 1312 30 30 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 1 1 32768 1 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 6 6 8192 6 6 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 33 33 4096 33 33 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 7 8 2048 4 4 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 57 60 1024 15 15 1
size-512(DMA) 0 0 512 0 0 1
size-512 16 24 512 2 3 1
size-256(DMA) 0 0 256 0 0 1
size-256 21 30 256 2 2 1
size-128(DMA) 2 30 128 1 1 1
size-128 462 510 128 17 17 1
size-64(DMA) 0 0 64 0 0 1
size-64 242 295 64 5 5 1
size-32(DMA) 2 113 32 1 1 1
size-32 429 1017 32 9 9 1
evol

2.while starting evolution (evolution-pine-importer looks through MANY
pine mboxes to see if it can import them, more or less like updatedb that
ckecks all the files in the disc, so it looks like something related to
reading lots of files)

slabinfo - version: 1.1
kmem_cache 59 68 112 2 2 1
ip_conntrack 7 33 352 3 3 1
ip_fib_hash 12 113 32 1 1 1
uhci_urb_priv 1 67 56 1 1 1
ip_mrt_cache 0 0 96 0 0 1
tcp_tw_bucket 1 40 96 1 1 1
tcp_bind_bucket 12 113 32 1 1 1
tcp_open_request 0 0 64 0 0 1
inet_peer_cache 1 59 64 1 1 1
ip_dst_cache 55 100 192 5 5 1
arp_cache 2 30 128 1 1 1
blkdev_requests 512 520 96 13 13 1
dnotify cache 0 0 20 0 0 1
file lock cache 4 42 92 1 1 1
fasync cache 1 202 16 1 1 1
uid_cache 7 113 32 1 1 1
skbuff_head_cache 161 240 160 10 10 1
sock 324 324 832 36 36 2
sigqueue 1024 1044 132 36 36 1
kiobuf 0 0 8768 0 0 4
cdev_cache 21 118 64 2 2 1
bdev_cache 6 59 64 1 1 1
mnt_cache 16 59 64 1 1 1
inode_cache 994 2178 416 242 242 1
dentry_cache 984 2730 128 91 91 1
dquot 0 0 128 0 0 1
filp 3116 3120 96 78 78 1
names_cache 1 4 4096 1 4 1
buffer_head 2229 3440 96 86 86 1
mm_struct 86 96 160 4 4 1
vm_area_struct 4929 4956 64 84 84 1
fs_cache 85 118 64 2 2 1
files_cache 85 90 416 10 10 1
signal_act 90 93 1312 31 31 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 1 2 32768 1 2 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 6 6 8192 6 6 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 37 38 4096 37 38 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 9 12 2048 5 6 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 57 60 1024 15 15 1
size-512(DMA) 0 0 512 0 0 1
size-512 20 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 39 60 256 4 4 1
size-128(DMA) 2 30 128 1 1 1
size-128 503 570 128 19 19 1
size-64(DMA) 0 0 64 0 0 1
size-64 248 295 64 5 5 1
size-32(DMA) 2 113 32 1 1 1
size-32 442 1017 32 9 9 1

BTW, It has taken me more than 10 minutes to write this email and
evolution hasn't yet started.

3.last slabinfo (the disk still "plays")

slabinfo - version: 1.1
kmem_cache 59 68 112 2 2 1
ip_conntrack 10 22 352 2 2 1
ip_fib_hash 12 113 32 1 1 1
uhci_urb_priv 1 67 56 1 1 1
ip_mrt_cache 0 0 96 0 0 1
tcp_tw_bucket 1 40 96 1 1 1
tcp_bind_bucket 11 113 32 1 1 1
tcp_open_request 0 0 64 0 0 1
inet_peer_cache 2 59 64 1 1 1
ip_dst_cache 45 100 192 5 5 1
arp_cache 2 30 128 1 1 1
blkdev_requests 512 520 96 13 13 1
dnotify cache 0 0 20 0 0 1
file lock cache 9 42 92 1 1 1
fasync cache 1 202 16 1 1 1
uid_cache 7 113 32 1 1 1
skbuff_head_cache 129 264 160 11 11 1
sock 344 351 832 39 39 2
sigqueue 433 435 132 15 15 1
kiobuf 0 0 8768 0 0 4
cdev_cache 21 118 64 2 2 1
bdev_cache 6 59 64 1 1 1
mnt_cache 16 59 64 1 1 1
inode_cache 1023 2178 416 242 242 1
dentry_cache 1012 2730 128 91 91 1
dquot 0 0 128 0 0 1
filp 3116 3120 96 78 78 1
names_cache 0 3 4096 0 3 1
buffer_head 2072 2520 96 63 63 1
mm_struct 84 96 160 4 4 1
vm_area_struct 5156 5251 64 89 89 1
fs_cache 83 118 64 2 2 1
files_cache 83 90 416 10 10 1
signal_act 88 93 1312 31 31 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 1 1 32768 1 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 6 6 8192 6 6 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 33 34 4096 33 34 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 7 10 2048 4 5 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 53 56 1024 14 14 1
size-512(DMA) 0 0 512 0 0 1
size-512 16 32 512 2 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 21 30 256 2 2 1
size-128(DMA) 2 30 128 1 1 1
size-128 461 510 128 17 17 1
size-64(DMA) 0 0 64 0 0 1
size-64 249 295 64 5 5 1
size-32(DMA) 2 113 32 1 1 1
size-32 440 1017 32 9 9 1

While the disc is reading:
   procs memory swap io system cpu
 r b w swpd free buff cache si so bi bo in cs us sy id
 3 11 0 2492 2800 436 8300 9 8 610 14 160 308 89 6 6
 2 7 0 2412 3568 408 8232 20 2 1538 4 222 264 76 9 15
 1 10 0 2664 2800 432 9128 3 30 1388 40 212 201 89 8 3
 2 7 0 2616 3020 492 9360 26 25 1457 25 204 198 84 6 10
 4 10 0 2648 2800 420 8780 8 0 1377 6 201 233 88 7 4
 1 10 0 2528 2800 440 8736 9 0 1248 0 220 280 82 6 12
 1 6 1 2520 2800 448 8688 1 3 1350 30 212 230 75 12 13

Pau

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sun Sep 30 2001 - 21:00:32 EST