■squidclientの統計情報を確認する。
※squid+squidguardの環境。
最初のヒット時だけ変に遅いのが気になった。。。
それ以外は快速。
■squidclientの導入。
※Windows版squidには「bin」配下にあり、同じように使える。
> bin\squidclient.exe
$ apt-cache search ^squidclient
squidclient - フル機能のウェブプロキシキャッシュ (HTTP プロキシ) - 制御ユーティリティ
■隠し属性の制御系のコマンドは使いません。
$ sudo squidclient mgr: | grep -v public
config Current Squid Configuration hidden
shutdown Shut Down the Squid Process hidden
reconfigure Reconfigure the Squid Process hidden
offline_toggle Toggle offline_mode setting hidden
■パブリック属性の統計コマンドを使います。
$ sudo squidclient mgr: | grep public | sed s/" *public"//g
mem Memory Utilization
cbdata Callback Data Registry Contents
events Event Queue
squidaio_counts Async IO Function Counters
coss COSS Stats
diskd DISKD Stats
ipcache IP Cache Stats and Contents
fqdncache FQDN Cache Stats and Contents
idns Internal DNS Statistics
url_rewriter URL Rewriter Stats
external_acl External ACL stats
http_headers HTTP Header Statistics
menu This Cachemanager Menu
info General Runtime Information
filedescriptors Process Filedescriptor Allocation
objects All Cache Objects
vm_objects In-Memory and In-Transit Objects
openfd_objects Objects with Swapout files open
pending_objects Objects being retreived from the network
client_objects Objects being sent to clients
io Server-side network read() size histograms
counters Traffic and Resource Counters
peer_select Peer Selection Algorithms
digest_stats Cache Digest and ICP blob
5min 5 Minute Average of Counters
60min 60 Minute Average of Counters
utilization Cache Utilization
histograms Full Histogram Counts
active_requests Client-side Active Requests
store_digest Store Digest
storedir Store Directory Stats
store_check_cachable_stats storeCheckCachable() Stats
store_io Store IO Interface Stats
pconn Persistent Connection Utilization Histograms
refresh Refresh Algorithm Statistics
delay Delay Pool Levels
forward Request Forwarding Statistics
client_list Cache Client List
asndb AS Number Database
server_list Peer Cache Statistics
non_peers List of Unknown sites sending ICP messages
■上記の細かいところは置いといて、システムメールします。
$ cat myscripts/squidsummary.sh
if [ "`id -u`" -ne "0" ];then
echo "Sorry,Not Permit User"
exit 1
fi
TODAY="`date '+%Y-%m-%d'`"
SQUIDSTAT="/var/log/squidsummary_$TODAY"
squidclient mgr:info > $SQUIDSTAT
squidclient mgr:60min >> $SQUIDSTAT
cat "$SQUIDSTAT" | \
mail -s "Squid Summary Report $TODAY" root
unset TODAY SQUIDSTAT
exit 0
■以下のように出ます。
ICPを使用していない(設定が0)であること、
キャッシュ率はもう少し調整の余地がありそうなこと、
やはり「DNS Lookups」が遅すぎること等が分かります。
4秒って何かのタイムアウトでしょうか。。。
Squid Object Cache: Version 2.7.STABLE9
Start Time: Sun, 26 Aug 2012 13:17:48 GMT
Current Time: Fri, 31 Aug 2012 12:51:22 GMT
Connection information for squid:
Number of clients accessing cache: 5
Number of HTTP requests received: 119340
Number of ICP messages received: 0
Number of ICP messages sent: 0
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: 16.6
Average ICP messages per minute since start: 0.0
Select loop called: 15281854 times, 28.165 ms avg
Cache information for squid:
Request Hit Ratios: 5min: 6.2%, 60min: 5.0%
Byte Hit Ratios: 5min: 0.0%, 60min: 0.0%
Request Memory Hit Ratios: 5min: 0.0%, 60min: 25.0%
Request Disk Hit Ratios: 5min: 0.0%, 60min: 0.0%
Storage Swap size: 92056 KB
Storage Mem size: 58280 KB
Mean Object Size: 11.92 KB
Requests given to unlinkd: 77973
Median Service Times (seconds) 5 min 60 min:
HTTP Requests (All): 0.00091 0.11465
Cache Misses: 0.00091 0.11465
Cache Hits: 0.00000 0.00000
Near Hits: 0.00000 5.34113
Not-Modified Replies: 0.00000 0.00000
DNS Lookups: 4.77162 4.77162
ICP Queries: 0.00000 0.00000
Resource usage for squid:
UP Time: 430414.434 seconds
CPU Time: 660.901 seconds
CPU Usage: 0.15%
CPU Usage, 5 minute avg: 1.03%
CPU Usage, 60 minute avg: 0.86%
Process Data Segment Size via sbrk(): 82664 KB
Maximum Resident Size: 333520 KB
Page faults with physical i/o: 34
Memory usage for squid via mallinfo():
Total space in arena: 82664 KB
Ordinary blocks: 80046 KB 898 blks
Small blocks: 0 KB 0 blks
Holding blocks: 396 KB 1 blks
Free Small blocks: 0 KB
Free Ordinary blocks: 2617 KB
Total in use: 80442 KB 97%
Total free: 2617 KB 3%
Total size: 83060 KB
Memory accounted for:
Total accounted: 71960 KB
memPoolAlloc calls: 30458307
memPoolFree calls: 30221180
File descriptor usage for squid:
Maximum number of file descriptors: 1024
Largest file desc currently in use: 54
Number of file desc currently in use: 28
Files queued for open: 0
Available number of file descriptors: 996
Reserved number of file descriptors: 100
Store Disk files open: 0
IO loop method: epoll
Internal Data Structures:
7899 StoreEntries
7496 StoreEntries with MemObjects
7493 Hot Object Cache Items
7724 on-disk objects
...
■60minは以下のように取得しなおし。
CPUの利用率も低く、スワップによる遅延も少ないのに、
やはりDNSがボトルネックになっています。
$ sudo squidclient mgr:60min | \
grep -v "0.000000" | sort -t\= -k 2 -nr | grep -v GMT | head -15 | sort
client_http.all_median_svc_time = 0.142521 seconds
client_http.kbytes_out = 221.445516/sec
client_http.nh_median_svc_time = 5.341132 seconds
cpu_time = 30.845927 seconds
cpu_usage = 0.856808%
dns.median_svc_time = 4.771619 seconds
select_fds = 259.283076/sec
select_loops = 253.184912/sec
server.all.kbytes_in = 221.367740/sec
server.http.kbytes_in = 221.255243/sec
syscalls.disk.writes = 0.187495/sec
syscalls.polls = 253.184912/sec
syscalls.sock.reads = 126.135107/sec
syscalls.sock.writes = 133.101025/sec
wall_time = 3600.099998 seconds
■なるほど、起動していない仮想のvyattaのDNSを見に行ってたわけですね。。。
$ sudo squidclient mgr:idns | grep -A 3 ^IP
IP ADDRESS
--------------- --------- ---------
192.168.72.213 1702 0
192.168.54.1 1702 1702
■「/etc/resolv.conf」を修正して解決です。
1時間置きに監視します。
$ sudo crontab -l | grep ^[0-9] | tail -1
55 * * * * /usr/local/sbin/myscripts/squidsummary.sh
