1 |
Duncan wrote: |
2 |
> |
3 |
> Pause every few seconds. I've seen two commonly reported solutions to |
4 |
> this -- two different problems resulting in the same symtoms -- and |
5 |
> another one less common in general but likely equally common on amd64. |
6 |
> |
7 |
> 1) DMA, or rather, the lack thereof, or interrupt sharing. Ensure that |
8 |
> you have the correct chipset drivers installed, both for whatever bus the |
9 |
> NIC is on, and for your hard drive interface (the PATA/SATA/SCSI chipset). |
10 |
> Often, generic or the wrong drivers will work in degraded mode, but DMA |
11 |
> won't be turned on, and performance will suffer. For your hard drives |
12 |
> (yes, wrong drivers there /can/ cause bottlenecks up the chain, including |
13 |
> onto the same bus as the NIC), use hdparm (or sdparm) to verify that both |
14 |
> the drive and the chipset agree that DMA is on. |
15 |
|
16 |
I am using the Promise EX8350 raid card and using there driver. |
17 |
|
18 |
> |
19 |
> Also check interrupt distribution and sharing. You may be able to change |
20 |
> these settings in the BIOS, or try switching some cards around |
21 |
> (particularly your NIC if it's not on-board built-in). |
22 |
|
23 |
This one I was thinking of trying the last. (Read below) |
24 |
|
25 |
> |
26 |
> 2) Do you have an "extra" network interface? I've seen this one on |
27 |
> MSWormOS and am unsure how Linux would even have the problem, unless you |
28 |
> have a spare interface (say a wireless card) that's running but not |
29 |
> connected. Anyway, the problem here is an interface set to DHCP mode that |
30 |
> can't get an IP because it's not connected, so of course can't see a |
31 |
> server. The system would pause every few minutes (every few seconds seems |
32 |
> a bit much, but...) and try to obtain an IP, stalling all I/O in the |
33 |
> process. (Again, the reason it's stalling all I/O is likely rooted in |
34 |
> misconfigured DMA or IRQs, but anyway...) |
35 |
|
36 |
All the network cards are set to static address. |
37 |
|
38 |
> |
39 |
> 3) Seen frequently on this list as a problem resulting in one or more of |
40 |
> a number of symtoms, including this one: APIC or ACPI problems. Take a |
41 |
> look at the kernel documentation under $kerneldir/Documentation/x86_64, |
42 |
> the apic and acpi boot options, and idle=poll as well. Note that those |
43 |
> needing these often find they don't after a mature BIOS is available and |
44 |
> flashed. idle=poll is particularly nasty, as it will cause your CPU to |
45 |
> work far harder and run far hotter than it ordinarily would, but some |
46 |
> early amd64 boards required it until an updated BIOS fixed the issue. I'm |
47 |
> not sure if that's still a problem with the newer boards or not. |
48 |
|
49 |
I am using the latest bios. So far I am not using any apic boot options. Will |
50 |
read up. |
51 |
|
52 |
> |
53 |
> Also note that for specific net chip, various configuration options may be |
54 |
> available. In particular, there's the Broadcom tigon3/tg3 kernel driver, |
55 |
> and then another (I forget what it's called) for the same net chip. Often |
56 |
> one will work well, the other won't work or will have issues. If that's |
57 |
> your chip, look it up and consider trying the other driver. |
58 |
|
59 |
Will check this. |
60 |
|
61 |
Thank you very much for replying and helping. Please read my observations below |
62 |
and if you have time please advice. Let me know if I still have to do the checks |
63 |
as mention above. |
64 |
|
65 |
Here are some specs and observations. |
66 |
|
67 |
Specs. |
68 |
cpu: AMD 64 3000+ |
69 |
drives: Sata drives |
70 |
mobo: MSI Neo 4 Platinum |
71 |
Ram: 2GB |
72 |
Raid: Promise EX8350 |
73 |
OS: Gentoo |
74 |
Kernel: 2.6.13 r3 |
75 |
Filesystem: XFS |
76 |
|
77 |
Observations. |
78 |
When I was using the xfs I had the problem of the pauses in the tcp transfers. |
79 |
These pauses, I am guessing is because of the pdflush that occurs around every |
80 |
20 seconds. There is only one pdflush every 20 seconds. This caused lose of |
81 |
video frames when we did the video recording. The video would have a jerk or |
82 |
jump when played back. |
83 |
|
84 |
I changed the dirty_ratio from the default of 40 to 5. This solved the problem. |
85 |
During the recording preview there seems to be some jerks but during the |
86 |
playback it seems to be ok. The pdflush happens more often and there is still |
87 |
one pdflush. I think the jerk that I see in the record preview coincides with |
88 |
the pdflush. More jerks in record preview but ok during playback. |
89 |
|
90 |
Next I formatted the drives to the ext3 filesystem. The dirty_ratio was set back |
91 |
to the default value of 40. When the video started to record there is no loss of |
92 |
frames and no jerks in the preview window. During the playback there is no loss. |
93 |
Everything is good. Check the top for pdflush and noticed that the there is two |
94 |
pdflush. The pdflushs is happening at every 5 seconds or so. Very fast compared |
95 |
to when I was using xfs. |
96 |
|
97 |
So the problem must be my xfs setting. For xfs I have only used the default |
98 |
setting. Or could there be a problem with xfs and AMD64 kernel? |
99 |
|
100 |
What must I do to get xfs to work? Or what am I doing wrong? |
101 |
|
102 |
Please comment and advise. |
103 |
|
104 |
Once again thank you very much for the advice. |
105 |
|
106 |
P.V.Anthony |
107 |
|
108 |
|
109 |
-- |
110 |
gentoo-amd64@g.o mailing list |