With squid 3.1.11 CPU usage of squid process is 100% during 10am to 10 pm...
I will try now with 2.7.Stable9. I just dont know what could be the problem.
On 23.3.11. 16.24, Marcus Kool wrote:
Zivanic Dejan wrote:
On 3/23/11 3:27 AM, Marcus Kool wrote:
Dejan,
Squid is known to be CPU
In my testing in the last couple of weeks, I've found that newer squid
versions take significantly more cpu than the older versions, translating
into significantly less capacity
I didn't test 2.7, but in my tests
3.0 4200 requests/sec
3.1.11 2100 requests/sec
3.2.0.5 1400 requests/sec (able
When squid is stop... cpu usage dont go over 5%.
Maybe my conf is problem, but I think it is not...
I have about 6k request per minute so I am confused with this poor
performance.
About 1k users access squid in peek with about 80Mbps traffic.
On Sat, 26 Mar 2011 14:06:10 -0700 (PDT),
the tests below are for requesting small pages (50 bytes)
when I ran the same tests with 100K pages the throughput dropped
down
3.0 90 requests/sec
3.1 60 requests/sec
so, at 6K requests/min (100 requests/sec), you could legitimatly have the
system maxed out (depending on your mixture of
On Tue, 22 Mar 2011, George Herbert wrote:
On Tue, Mar 22, 2011 at 7:27 PM, Marcus Kool
marcus.k...@urlfilterdb.com wrote:
Dejan,
Squid is known to be CPU bound under heavy load and the
Quad core running at 1.6 GHz in not the fastest.
A 3.2 GHz dual core will give you double speed.
Second
Il 22/03/2011 20.30, Dejan Zivanic ha scritto:
acl kroz-adsl url_regex -i /etc/adsl
I'm no squid expert, but url_regex are known to be quit cpu-hungry. How
many entries does /etc/adsl have ?
Probably moving regex matching to an external daemon would allow to make
better use of the cpu
Regards,
we have heavy load (over 6k requests per minute) intercepting squid
loading about 70-80Mbps traffic.
I have notices that CPU usage of squid process never goes down from 50%
and usually goes up to over 90%.
We plan to upgrade to 120Mbps link and this can be major problem if we
On Tue, 22 Mar 2011 20:30:11 +0100, Dejan Zivanic wrote:
Regards,
we have heavy load (over 6k requests per minute) intercepting squid
loading about 70-80Mbps traffic.
I have notices that CPU usage of squid process never goes down from
50% and usually goes up to over 90%.
We plan to upgrade to
Amos,
He said over 6k requests per minute, not 6k per second.
Can you paste '/etc/adsl' or at least comment that acl and try again?
Dejan,
Squid is known to be CPU bound under heavy load and the
Quad core running at 1.6 GHz in not the fastest.
A 3.2 GHz dual core will give you double speed.
The config parameter minimum_object_size 10 KB
prevents that objects smaller than 10 KB are not written to disk.
I am curious to know
On Tue, Mar 22, 2011 at 7:27 PM, Marcus Kool
marcus.k...@urlfilterdb.com wrote:
Dejan,
Squid is known to be CPU bound under heavy load and the
Quad core running at 1.6 GHz in not the fastest.
A 3.2 GHz dual core will give you double speed.
Second this. CPU speed - perf wasn't quite linear
Felix New wrote:
Hi,all:
Our squid box's performance is not very good(4 reqs/min), cpu
load is very high, and the traffic can't great bigger than
100MBits/s.(Gbits/s network).
We use aufs scheme now.We have a test for coss, and the result is
not very good also(cpu overload also).
Hi,all:
Our squid box's performance is not very good(4 reqs/min), cpu
load is very high, and the traffic can't great bigger than
100MBits/s.(Gbits/s network).
We use aufs scheme now.We have a test for coss, and the result is
not very good also(cpu overload also).
can we optimize
On Thu, Mar 22, 2007, Felix New wrote:
Hi,all:
Our squid box's performance is not very good(4 reqs/min), cpu
load is very high, and the traffic can't great bigger than
100MBits/s.(Gbits/s network).
We use aufs scheme now.We have a test for coss, and the result is
not very
2007/3/22, Adrian Chadd [EMAIL PROTECTED]:
Do some profiling to find out whether you're lacking CPU or IO grunt.
I started optimising the Squid-2 codebase to remove a lot of the CPU
hogs but its a big job and noone seemed willing to sponsor any work.
Try out the Squid-2-HEAD codetree and see
Hello.
I did manage to lower CPU by setting squid.conf timeouts. From average
load 1.0 to load .0.7
Which is much better. (maybe not for persistent connections, etc.).
the epool... I did notice some info. But this would be my last
resort/hope. (I would prefer stable enviroment)
I plan to
Hello.
I am new to this list and new to squid under heavy load...
my problem is that squid spends 70% of CPU time in system.
the number is from top. ( 27.9% user, 66.8% system) and squid is the
only thing running.
it has open around 2100 descriptors.
it is handlig around 300req/sec
most of the
On Monday 12 September 2005 12:16, Michal Mihalik wrote:
Hello.
I am new to this list and new to squid under heavy load...
my problem is that squid spends 70% of CPU time in system.
the number is from top. ( 27.9% user, 66.8% system) and squid is the
only thing running.
You need to find
Hi.
thanks. for realy quick answer.
One important information which is missing here. It is running in
Accelereation mode.
Denis Vlasenko wrote:
On Monday 12 September 2005 12:16, Michal Mihalik wrote:
Hello.
I am new to this list and new to squid under heavy load...
my problem is that
-Original Message-
From: Denis Vlasenko [mailto:[EMAIL PROTECTED]
On Monday 12 September 2005 12:16, Michal Mihalik wrote:
system SPEC 2x.
I didn't understand this part.
I think he means he has 2 of these machines
Debian stable. 2.4 kernel.
Squid Cache: Version
Hi,
-Original Message-
From: Ow Mun Heng [mailto:[EMAIL PROTECTED]
Sent: Monday, September 12, 2005 12:01 PM
To: Denis Vlasenko; squid-users@squid-cache.org
Cc: Michal Mihalik
Subject: RE: [squid-users] Heavy LOAD on squid - system time
-Original Message-
From
On Mon, 12 Sep 2005, Michal Mihalik wrote:
A big quietion is how much will help to move to 2.6.x kernel.
and if the answer is 10%, the next question is I NEED MORE :-) how... please.
The big win for 2.6 would probably be that it has epoll support, allowing
you to experiment with using epoll
Hi ...
We are using Squid 2.5, and on peak times we have 100 users connected
Browsing ... And all goes Well , but a friend of mine in other
more big campus of my university, try squid but there are 700 users
connected on peak times, browsing ...
And after a while squid becomes very slow, and
Some information.
Number of HTTP requests received: 648417
Number of ICP messages received:33035
Number of ICP messages sent:35967
Number of queued ICP replies: 0
Request failure ratio: 0.00%
Average HTTP requests per
fre 2003-02-21 klockan 09.24 skrev kris:
I use disk and have 3 disks, load average isn't probably good measure but
system load h login procedure takes some time and doing
something takes some time and sometime reloading squid crashes the
squid process :)))
I change
What kind of raid are you using? (raid level)
On how many physical disks?
And when did your cache become filled? There is a significant
difference in performance between an empty cache being filled and a
cache which is full and being recycled..
Regards
Henrik
fre 2003-02-21 klockan 13.09
What kind of raid are you using? (raid level)
no raid level, only the controller each disk separatelly
On how many physical disks?
3, one system and two adre cache disks
And when did your cache become filled? There is a significant
difference in performance between an empty cache
fre 2003-02-21 klockan 14.00 skrev kris:
no raid level, only the controller each disk separatelly
Good.
3, one system and two adre cache disks
2 cache disks is not sufficient for the load you are trying to push I
think. At least not unless these drives are solid state drives with no
moving
I have a squid machine based on 1 Xeon 2.8GHz and SCSI discs, i have about
260 req/s and very high load average (about 4 to 5)
My squid is Version 2.5.STABLE1-20030114 running on RH8.0 SCSI U160 Xeon
2.8GHZ
Is that load normal ? I think that it's to high
What's the problem ?
I guess too
I have a squid machine based on 1 Xeon 2.8GHz and SCSI discs, i have
about 260 req/s and very high load average (about 4 to 5)
My squid is Version 2.5.STABLE1-20030114 running on RH8.0 SCSI U160
Xeon 2.8GHZ
Is that load normal ? I think that it's to high
What's the problem ?
I guess
Are you using aufs?
If you are then the load average is not a very good measure of load..
better to measure the amount of idle CPU time, system throughput, and
response times.
It is possible that the aufs performance fixes simply have made your
server work even better, giving a higher load
31 matches
Mail list logo