[squid-users] RE: TCP response time | Proxy efficiency | Paging .. pending questions

>
> Hello all,
>
> when looking on the good working of one of my squid proxies, I see
some values I think somewhat relevant for increasing the cache_mem.
>
> next to my investigations I can see this machine receives some
130req/s, sends some 80 req/s, caches some 1.2 million objects on disk.
this is the very ba sic part of it.
>
> $ squidclient mgr:info
> Internal Data Structures:
> 1182755 StoreEntries
> 70431 StoreEntries with MemObjects
> 70365 Hot Object Cache Items
> 1174685 on-disk objects
>
> $ squidclient mgr:5min |egrep 'http|fault'
> client_http.requests = 127.388482/sec
> client_http.hits = 45.899453/sec
> client_http.errors = 0.000000/sec
> client_http.kbytes_in = 95.962190/sec
> client_http.kbytes_out = 1820.344974/sec
> client_http.all_median_svc_time = 0.021898 seconds
> client_http.miss_median_svc_time = 0.042766 seconds
> client_http.nm_median_svc_time = 0.001789 seconds
> client_http.nh_median_svc_time = 0.020695 seconds
> client_http.hit_median_svc_time = 0.004626 seconds
> server.http.requests = 78.912393/sec
> server.http.errors = 0.000000/sec
> server.http.kbytes_in = 1285.901343/sec
> server.http.kbytes_out = 66.245877/sec
> page_faults = 0.043333/sec
>
> we can also see a correct hit (36%) and byte (20%) rate on this server
but ...
>
> * what does it mean "Proxy efficiency 46.01" ??
> * and what does it mean "Average speed increase 24.27% " ??
> * what does it mean "TCP response time of 100%% requests" ..
> ? the maximum time taken to make the SYN SYN/ACK ACK when
asking for a socket on the parent cahe
> ? the maximum time taken for getting an object from the
parent cache
> ???
>
> Summary
> Calamaris statistics
>
> lines parsed: lines 15855657
> invalid lines: lines 0
> parse time: sec 4145
> parse speed: lines/sec 3825
>
> Proxy statistics
>
> Total amount: requests 15855657
> Total Bandwidth: Byte 165G
> Proxy efficiency (HIT [kB/sec] / DIRECT [kB/sec]): factor 46.01
> Average speed increase: % 24.27
> TCP response time of 100%% requests: msec 1900
>
> Cache statistics
>
> Total amount cached: requests 5627335
> Request hit rate: % 35.49
> Bandwidth savings: Byte 33781M
> Bandwidth savings in Percent (Byte hit rate): % 19.96
> Average cached object size: Byte 6294
> Average direct object size: Byte 13885
> Average object size: Byte 11191
>
> the machine squid is running on is a Sun Solaris 8 V210 with 2Gb
memory.
>
> $ prtconf
> System Configuration: Sun Microsystems sun4u
> Memory size: 2048 Megabytes
> System Peripherals (Software Nodes):
> SUNW,Sun-Fire-V210
>
> concerning the memory usage this becomes much more unclear for me ...
if I good understand I see the process size is some 700mb and get some
600mb allocat ed. when going deeper in the vmstat stats pi and po column
respectively give 825 and 236 at some time ...
>
>
> * is this not some too big value ??
> * cache_mem is currently defined at 400mb, don't you think increasing
this value to a bigger value (let's say some 500-600 mb) would not be
better ??
>
>
> $ squidclient mgr:info
> Squid Object Cache: Version 2.7.STABLE4
> Start Time: Mon, 17 Aug 2009 02:00:38 GMT
> Current Time: Thu, 20 Aug 2009 10:48:06 GMT
> Connection information for squid:
> Number of clients accessing cache: 2264
> Number of HTTP requests received: 11355087
> Number of ICP messages received: 11553839
> Number of ICP messages sent: 11564676
> Number of queued ICP replies: 0
> Number of HTCP messages received: 0
> Number of HTCP messages sent: 0
> Request failure ratio: 0.00
> Average HTTP requests per minute since start: 2342.5
> Average ICP messages per minute since start: 4769.2
> Select loop called: 94346313 times, 3.083 ms avg
> Cache information for squid:
> Request Hit Ratios: 5min: 38.8%, 60min: 30.9%
> Byte Hit Ratios: 5min: 21.8%, 60min: 29.7%
> Request Memory Hit Ratios: 5min: 27.4%, 60min: 27.1%
> Request Disk Hit Ratios: 5min: 23.4%, 60min: 18.0%
> Storage Swap size: 29196377 KB
> Storage Mem size: 409632 KB
> Mean Object Size: 24.85 KB
> Requests given to unlinkd: 114
> Median Service Times (seconds) 5 min 60 min:
> HTTP Requests (All): 0.02190 0.01035
> Cache Misses: 0.04047 0.03622
> Cache Hits: 0.00463 0.00379
> Near Hits: 0.01955 0.02317
> Not-Modified Replies: 0.00179 0.00179
> DNS Lookups: 0.00190 0.00190
> ICP Queries: 0.00221 0.00221
> Resource usage for squid:
> UP Time: 290848.297 seconds
> CPU Time: 26579.690 seconds
> CPU Usage: 9.14%
> CPU Usage, 5 minute avg: 30.92%
> CPU Usage, 60 minute avg: 32.65%
> Process Data Segment Size via sbrk(): 693792 KB
> Maximum Resident Size: 0 KB
> Page faults with physical i/o: 19875
> Memory accounted for:
> Total accounted: 603121 KB
> memPoolAlloc calls: 1390282241
> memPoolFree calls: 1384735607
> File descriptor usage for squid:
> Maximum number of file descriptors: 8192
> Largest file desc currently in use: 801
> Number of file desc currently in use: 697
> Files queued for open: 0
> Available number of file descriptors: 7495
> Reserved number of file descriptors: 100
> Store Disk files open: 0
> IO loop method: poll
>
> $ vmstat 5 5
> procs memory page disk faults
cpu
> r b w swap free re mf pi po fr de sr m0 m1 m3 m4 in sy cs
us sy id
> 0 0 0 2614824 634360 62 277 76 29 29 0 0 3 0 3 5 4294967261 276
176 10 9 81
> 0 0 0 2442720 414768 145 128 568 236 236 0 0 0 0 0 60 3418 15771 2959
31 30 39
> 2 1 0 2442304 418104 296 851 739 75 75 0 0 27 0 1 20 2862 58159 2199
41 45 14
> 0 0 0 2442720 423744 229 0 825 180 180 0 0 0 0 0 8 2749 36179 2911
36 28 36
> 0 0 0 2442720 428832 146 0 452 96 96 0 0 0 0 1 4 3825 13153 2585
26 30 45
>
> $ vmstat -i
> interrupt total rate
> --------------------------------
> clock 1582676343 100
> --------------------------------
> Total 1582676343 100
>
>
> $ vmstat -s
> 0 swap ins
> 0 swap outs
> 0 pages swapped in
> 0 pages swapped out
> 119188729 total address trans. faults taken
> 63414018 page ins
> 33261045 page outs
> 150679474 pages paged in
> 57581487 pages paged out
> 985466858 total reclaims
> 985358256 reclaims from free list
> 0 micro (hat) faults
> 119188729 minor (as) faults
> 34878720 major faults
> 1518806042 copy-on-write faults
> 655083663 zero fill page faults
> 187523 pages examined by the clock daemon
> 0 revolutions of the clock hand
> 57477049 pages freed by the clock daemon
> 47526220 forks
> 149015 vforks
> 49086268 execs
> 3171912755 cpu context switches
> 1401743919 device interrupts
> 3042943801 traps
> 2429831033 system calls
> 1065243178 total name lookups (cache hits 87%)
> 151392974 user cpu
> 148588621 system cpu
> 1165931599 idle cpu
> 116763714 wait cpu
>
> many thks to help me and do not hesitate to give me any comment about
the files give above.
>

any news ??

> Vincent.
>
-----------------------------------------------------------------
ATTENTION:
The information in this electronic mail message is private and
confidential, and only intended for the addressee. Should you
receive this message by mistake, you are hereby notified that
any disclosure, reproduction, distribution or use of this
message is strictly prohibited. Please inform the sender by
reply transmission and delete the message without copying or
opening it.

Messages and attachments are scanned for all viruses known.
If this message contains password-protected attachments, the
files have NOT been scanned for viruses by the ING mail domain.
Always scan attachments before opening them.
-----------------------------------------------------------------
Received on Thu Aug 27 2009 - 12:24:30 MDT