1 |
On one of my machines doing a portage update is slow, horribly, |
2 |
painfully, slow. The rsync itself is fine, but after that, the |
3 |
"updating portage cache" is slow, but if there are any packages to move |
4 |
around with the "global updates" (fixpackages) it is amazingly slow. |
5 |
Rough calculations gives a minute to two minutes between each "*" that |
6 |
indicate some data written on the hard drive. Slow like I'm using a 286 |
7 |
to copy gigs of data around. |
8 |
|
9 |
The update I'm doing right now has taken 10 minutes or more, and it's |
10 |
still just updating one package! |
11 |
|
12 |
My system is no speed demon, but it's not a slouch either. Celeron 533, |
13 |
128mb ram, with the root partition being RAID0, 2 SCSI 2G drives |
14 |
attached to an old Adaptec 2940 Ultra SCSI adapter. |
15 |
|
16 |
A check of hdparm shows that I'm only getting around 10mb/s transfer |
17 |
with the RAID, and individually about 5mb/s, but that is typical I |
18 |
think of the fact that it's older hardware, but still... |
19 |
|
20 |
The rest of the system "feel"s just fine. I can copy a 28mb kernel |
21 |
tarball in 3.55s within the same drive, and 3.65 to the IDE drive I have |
22 |
in the system for storage. Other portage operations, emerging packages |
23 |
etc also feel fine, and in line with the speed of the system. |
24 |
|
25 |
The system itself is a webserver that has apache, postfix, squid, samba |
26 |
and not much else running (squid is caching onto the root drive BTW). |
27 |
If it was a windows system I'd be checking for adware running in the |
28 |
background or defragging my hard drive, but it's ext3 :) The system |
29 |
has been up for 60 days, so it hasn't been fscked in a while, but based |
30 |
on what I've seen of linux filesystems, it's not a defrag issue :) |
31 |
|
32 |
Any thoughts? Specs and benchmarks below. If anyone can suggest a |
33 |
way to speed this up, or a suggestion as to why it would be so slow, |
34 |
I'd be most appreciated. |
35 |
|
36 |
hdparm -tT on /dev/root (raid0) /dev/sda1 /dev/sdb1 |
37 |
--------------------------------------------------- |
38 |
/dev/root: |
39 |
Timing buffer-cache reads: 160 MB in 2.00 seconds = 80.00 MB/sec |
40 |
Timing buffered disk reads: 32 MB in 3.18 seconds = 10.06 MB/sec |
41 |
|
42 |
/dev/sda1: |
43 |
Timing buffer-cache reads: 184 MB in 2.00 seconds = 92.00 MB/sec |
44 |
Timing buffered disk reads: 16 MB in 3.13 seconds = 5.11 MB/sec |
45 |
|
46 |
/dev/sdb1: |
47 |
Timing buffer-cache reads: 184 MB in 2.01 seconds = 91.54 MB/sec |
48 |
Timing buffered disk reads: 14 MB in 3.00 seconds = 4.67 MB/sec |
49 |
|
50 |
dmesg |
51 |
----- |
52 |
scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.8 |
53 |
<Adaptec 2940 Ultra SCSI adapter> |
54 |
aic7880: Ultra Wide Channel A, SCSI Id=7, 16/253 SCBs |
55 |
|
56 |
Vendor: SEAGATE Model: ST32155W Rev: 0528 |
57 |
Type: Direct-Access ANSI SCSI revision: 02 |
58 |
(scsi0:A:1): 40.000MB/s transfers (20.000MHz, offset 8, 16bit) |
59 |
Vendor: Quantum Model: XP32150W Rev: L915 |
60 |
Type: Direct-Access ANSI SCSI revision: 02 |
61 |
(scsi0:A:3): 20.000MB/s transfers (10.000MHz, offset 8, 16bit) |
62 |
scsi0:A:1:0: Tagged Queuing enabled. Depth 253 |
63 |
scsi0:A:3:0: Tagged Queuing enabled. Depth 253 |
64 |
Attached scsi disk sda at scsi0, channel 0, id 1, lun 0 |
65 |
Attached scsi disk sdb at scsi0, channel 0, id 3, lun 0 |
66 |
SCSI device sda: 4197405 512-byte hdwr sectors (2149 MB) |
67 |
/dev/scsi/host0/bus0/target1/lun0: p1 p2 |
68 |
SCSI device sdb: 4199760 512-byte hdwr sectors (2150 MB) |
69 |
/dev/scsi/host0/bus0/target3/lun0: p1 p2 |
70 |
|
71 |
|
72 |
TIA |
73 |
|
74 |
|
75 |
-- |
76 |
Alan <alan@×××××.org> - http://arcterex.net |
77 |
-------------------------------------------------------------------- |
78 |
"There are only 3 real sports: bull-fighting, car racing and mountain |
79 |
climbing. All the others are mere games." -- Hemingway |
80 |
|
81 |
-- |
82 |
gentoo-dev@g.o mailing list |