tokyo tyrant采用的是从机向主机拉式的主从同步策略,并且有一个限制,一个从库只能从一个主库同步数据。另外tokyo tyrant在写操作的时候都会加锁。这样对同一个key的写操作就会被顺序执行,不会出现并发操作的情况。且主辅库均可进行读写操作。
下面是几种同步策略:
1.
2.
3.
DevOps
tokyo tyrant采用的是从机向主机拉式的主从同步策略,并且有一个限制,一个从库只能从一个主库同步数据。另外tokyo tyrant在写操作的时候都会加锁。这样对同一个key的写操作就会被顺序执行,不会出现并发操作的情况。且主辅库均可进行读写操作。
下面是几种同步策略:
1.
2.
3.
如下的遍历方式效率是比较低的:
1. 因为是用的迭代的方式,所以内存使用很少
2. 迭代的过程如下,效率是非常地的:
sendto(3, "\310Q", 2, 0, NULL, 0) = 2 (continue)
recvfrom(3, "\0\0\0\0CTGT-MTI5NjI0OTk5MA==-130492"…, 65536, 0, NULL, NULL) = 72 (接收key)
sendto(3, "\3100\0\0\0CTGT-MTI5NjI0OTk5MA==-13049"…, 73, 0, NULL, 0) = 73 (get key)
recvfrom(3, "\0\0\0\0L{\"uid\":\"1296249990\",\"et\":13"…, 65536, 0, NULL, NULL) = 81 (接收value)
测试脚本:
以前写过一个multi_http()的函数,就是异步地做http请求,后来再次看那段代码的时候,发现不少问题:
1. 我只在发送、接受数据的时候使用了异步,connect的时候还是同步的
2. 由于我发送数据的时候使用的是http1.0,所以接受数据的时候靠feof()判断结束就很方便了;如果使用http1.1,而且connection:keep-alive; 那么就不是那么简单了
如果需要异步并发,建议使用curl,今天看了一下,curl在异步并发请求的时候,connect、send、recv都是异步的。测试代码:
使用strace观察一下:
strace -tt php multi_curl.php
connect(3,{sa_family=AF_INET,sin_port=htons(80), sin_addr=inet_addr("66.147.244.189")}, 16) = -1 EINPROGRESS (Operation now in progress)
poll([{fd=3, events=POLLOUT}], 1, 0) = 0
虽然是异步,这里还是立即检查了一下是否已经连接成功;但是这种检查也是非阻塞的(看poll的第三个参数)。如果connect连接的是本地端口,poll检查的时候连接就已经是成功的了。 如(10.55.38.14是本机):
connect(4, {sa_family=AF_INET, sin_port=htons(80), sin_addr=inet_addr("10.55.38.14")}, 16) = -1 EINPROGRESS (Operation now in progress)
poll([{fd=4, events=POLLOUT, revents=POLLOUT}], 1, 0) = 1
——————————————
下面这篇文章写的并不好,仅作学习之用:
这里关键是socket_select的用法了;
另:
1. 当阻塞方式connect的时候,设置连接超时时间是通过设置SO_SNDTIMEO来实现的,如:
socket_set_option($socket, SOL_SOCKET, SO_SNDTIMEO, array(‘sec’ => 3, ‘usec’ => 0));
设置超时时间为3s;
2. 对于非阻塞方式的connect,连接的超时时间不是在socket上设置的,而是自己控制循环的时间
socket_shutdown函数可以关闭一个socket的读或写,这种操作映射到tcp协议中又是什么情况呢,简单测试了一下:
1. 如果关闭读操作,则协议上没有任何动作,对方发来的数据网卡上还是接受的
2. 如果关闭写操作,则协议上表现为关闭连接,即: 发送finish报文
关键字: 配置文件、日志、默认值
我们总期望程序功能很强大,但是有些功能不是适合任何环境的、任何时候的,这时候我们就可以利用配置来使得灵活。
比如:
我想方便地看到线上程序的一些调试信息,但是一般情况下我并不需要这些信息,只有在调试线上bug的时候才需要,于是我们就可以在程序中添加这些调试信息的语句,然后通过开关来控制。
如果我们没有权限登录线上提供服务的机器,则调试信息就可能需要输出到自己可以登录的某远程server上了,然而,很多提供服务的机器都往一个机器上打log,如果没有控制,这个log server估计会吃不消了,于是我们就可能需要配置一些有百分之多少的请求是需要打log的,这里就又用到了配置。
另外,还有日志的级别也是可以配置的。
所以,写一套系统,一定要有:
1. 配置文件类
什么配置信息都可以从通过配置文件类来获取;或许你不愿意配置这些东西,你们你设置默认值就行了
2. 日志类
日志有级别,也有标签;我可以输出某种级别的日志,也可以只输出含有某标签的日志; 或许你不想输出这些东西,你们你默认不输出就行了
include("jQuery");
// 在 IE 中无效:
$("<input>").attr("type", "checkbox");
// 在 IE 中有效:
$("<input type=’checkbox’>");
下载地址: http://www.thekelleys.org.uk/dnsmasq/dnsmasq-2.57.tar.gz
帮助文档: https://help.ubuntu.com/community/Dnsmasq
配置文件:/etc/dnsmasq.conf
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 |
# Configuration file for <a title="dnsmasq" href="http://htyp.org/dnsmasq">dnsmasq</a>. # # Format is one option per line, legal options are the same # as the long options legal on the command line. See # "/usr/sbin/dnsmasq --help" or "man 8 dnsmasq" for details. # The following two options make you a better netizen, since they # tell dnsmasq to filter out queries which the public DNS cannot # answer, and which load the servers (especially the root servers) # uneccessarily. If you have a dial-on-demand link they also stop # these requests from bringing up the link uneccessarily. # Never forward plain names (without a dot or domain part) domain-needed # Never forward addresses in the non-routed address spaces. bogus-priv # Uncomment this to filter useless windows-originated DNS requests # which can trigger dial-on-demand links needlessly. # Note that (amongst other things) this blocks all SRV requests, # so don't use it if you use eg Kerberos. # This option only affects forwarding, SRV records originating for # dnsmasq (via srv-host= lines) are not suppressed by it. #filterwin2k # Change this line if you want dns to get its upstream servers from # somewhere other that /etc/<a title="resolv.conf" href="http://htyp.org/resolv.conf">resolv.conf</a> #resolv-file= resolv-file=/etc/dnsmasqupstreamservers # By default, dnsmasq will send queries to any of the upstream # servers it knows about and tries to favour servers to are known # to be up. Uncommenting this forces dnsmasq to try each query # with each server strictly in the order they appear in # /etc/resolv.conf #strict-order # If you don't want dnsmasq to read /etc/resolv.conf or any other # file, getting its servers from this file instead (see below), then # uncomment this #no-resolv # If you don't want dnsmasq to poll /etc/resolv.conf or other resolv # files for changes and re-read them then uncomment this. #no-poll # Add other name servers here, with domain specs if they are for # non-public domains. #server=/localnet/192.168.0.1 #server=166.70.63.3 # Add local-only domains here, queries in these domains are answered # from /etc/hosts or DHCP only. #local=/localnet/ # Add domains which you want to force to an IP address here. # The example below send any host in doubleclick.net to a local # webserver. #address=/doubleclick.net/127.0.0.1 # If you want dnsmasq to change uid and gid to something other # than the default, edit the following lines. #user= #group= # If you want dnsmasq to listen for DHCP and DNS requests only on # specified interfaces (and the loopback) give the name of the # interface (eg eth0) here. # Repeat the line for more than one interface. #interface= # Or you can specify which interface _not_ to listen on #except-interface= # Or which to listen on by address (remember to include 127.0.0.1 if # you use this.) #listen-address= # If you want dnsmasq to provide only DNS service on an interface, # configure it as shown above, and then use the following line to # disable DHCP on it. #no-dhcp-interface= # On systems which support it, dnsmasq binds the wildcard address, # even when it is listening on only some interfaces. It then discards # requests that it shouldn't reply to. This has the advantage of # working even when interfaces come and go and change address. If you # want dnsmasq to really bind only the interfaces it is listening on, # uncomment this option. About the only time you may need this is when # running another nameserver on the same machine. #bind-interfaces # If you don't want dnsmasq to read /etc/hosts, uncomment the # following line. #no-hosts # or if you want it to read another file, as well as /etc/hosts, use # this. #addn-hosts=/etc/banner_add_hosts addn-hosts=/etc/dnsmasqhosts # Set this (and domain: see below) if you want to have a domain # automatically added to simple names in a hosts-file. #expand-hosts # Set the domain for dnsmasq. this is optional, but if it is set, it # does the following things. # 1) Allows DHCP hosts to have fully qualified domain names, as long # as the domain part matches this setting. # 2) Sets the "domain" DHCP option thereby potentially setting the # domain of all systems configured by DHCP # 3) Provides the domain part for "expand-hosts" #domain=thekelleys.org.uk # Uncomment this to enable the integrated DHCP server, you need # to supply the range of addresses available for lease and optionally # a lease time. If you have more than one network, you will need to # repeat this for each network on which you want to supply DHCP # service. dhcp-range=192.168.0.50,192.168.0.150,12h # This is an example of a DHCP range where the netmask is given. This # is needed for networks we reach the dnsmasq DHCP server via a relay # agent. If you don't know what a DHCP relay agent is, you probably # don't need to worry about this. #dhcp-range=192.168.0.50,192.168.0.150,255.255.255.0,12h # This is an example of a DHCP range with a network-id, so that # some DHCP options may be set only for this network. #dhcp-range=red,192.168.0.50,192.168.0.150 # Supply parameters for specified hosts using DHCP. There are lots # of valid alternatives, so we will give examples of each. Note that # IP addresses DO NOT have to be in the range given above, they just # need to be on the same network. The order of the parameters in these # do not matter, it's permissble to give name,adddress and MAC in any order # Always allocate the host with ethernet address 11:22:33:44:55:66 # The IP address 192.168.0.60 #dhcp-host=11:22:33:44:55:66,192.168.0.60 # Always set the name of the host with hardware address # 11:22:33:44:55:66 to be "fred" #dhcp-host=11:22:33:44:55:66,fred # Always give the host with ethernet address 11:22:33:44:55:66 # the name fred and IP address 192.168.0.60 and lease time 45 minutes #dhcp-host=11:22:33:44:55:66,fred,192.168.0.60,45m dhcp-host=00:0e:2e:0d:44:be,Bunsen,192.168.0.106 # Give the machine which says it's name is "bert" IP address # 192.168.0.70 and an infinite lease #dhcp-host=bert,192.168.0.70,infinite # Always give the host with client identifier 01:02:02:04 # the IP address 192.168.0.60 #dhcp-host=id:01:02:02:04,192.168.0.60 # Always give the host with client identifier "marjorie" # the IP address 192.168.0.60 #dhcp-host=id:marjorie,192.168.0.60 # Enable the address given for "judge" in /etc/hosts # to be given to a machine presenting the name "judge" when # it asks for a DHCP lease. #dhcp-host=judge # Never offer DHCP service to a machine whose ethernet # address is 11:22:33:44:55:66 #dhcp-host=11:22:33:44:55:66,ignore # Ignore any client-id presented by the machine with ethernet # address 11:22:33:44:55:66. This is useful to prevent a machine # being treated differently when running under different OS's or # between PXE boot and OS boot. #dhcp-host=11:22:33:44:55:66,id:* # Send extra options which are tagged as "red" to # the machine with ethernet address 11:22:33:44:55:66 #dhcp-host=11:22:33:44:55:66,net:red # Send extra options which are tagged as "red" to # any machine with ethernet address starting 11:22:33: #dhcp-host=11:22:33:*:*:*,net:red # Send extra options which are tagged as "red" to any machine whose # DHCP vendorclass string includes the substring "Linux" #dhcp-vendorclass=red,Linux # Send extra options which are tagged as "red" to any machine one # of whose DHCP userclass strings includes the substring "accounts" #dhcp-userclass=red,accounts # If this line is uncommented, dnsmasq will read /etc/ethers and act # on the ethernet-address/IP pairs found there just as if they had # been given as --dhcp-host options. Useful if you keep # MAC-address/host mappings there for other purposes. #read-ethers # Send options to hosts which ask for a DHCP lease. # See <a href="http://tools.ietf.org/html/rfc2132">RFC 2132</a> for details of available options. # Note that all the common settings, such as netmask and # broadcast address, DNS server and default route, are given # sane defaults by dnsmasq. You very likely will not need any # any dhcp-options. If you use Windows clients and Samba, there # are some options which are recommended, they are detailed at the # end of this section. # For reference, the common options are: # subnet mask - 1 # default router - 3 # DNS server - 6 # broadcast address - 28 # Set the NTP time server addresses to 192.168.0.4 and 10.10.0.5 #dhcp-option=42,192.168.0.4,10.10.0.5 dhcp-option=3,192.168.0.1 # Set the NTP time server address to be the same machine as # is running dnsmasq #dhcp-option=42,0.0.0.0 # Set the NIS domain name to "welly" #dhcp-option=40,welly # Set the default time-to-live to 50 #dhcp-option=23,50 # Set the "all subnets are local" flag #dhcp-option=27,1 # Send the etherboot magic flag and then etherboot options (a string). #dhcp-option=128,e4:45:74:68:00:00 #dhcp-option=129,NIC=eepro100 # Specify an option which will only be sent to the "red" network # (see dhcp-range for the declaration of the "red" network) #dhcp-option=red,42,192.168.1.1 # The following DHCP options set up dnsmasq in the same way as is specified # for the ISC dhcpcd in # <a href="http://www.samba.org/samba/ftp/docs/textdocs/DHCP-Server-Configuration.txt" rel="nofollow">http://www.samba.org/samba/ftp/docs/textdocs/DHCP-Server-Configuration.txt</a> # adapted for a typical dnsmasq installation where the host running # dnsmasq is also the host running samba. # you may want to uncomment them if you use Windows clients and Samba. #dhcp-option=19,0 # option ip-forwarding off #dhcp-option=44,0.0.0.0 # set netbios-over-TCP/IP nameserver(s) aka WINS server(s) #dhcp-option=45,0.0.0.0 # netbios datagram distribution server #dhcp-option=46,8 # netbios node type #dhcp-option=47 # empty netbios scope. # Send RFC-3397 DNS domain search DHCP option. WARNING: Your DHCP client # probably doesn't support this...... #dhcp-option=119,eng.apple.com,marketing.apple.com # Send encapsulated vendor-class specific options. The vendor-class # is sent as DHCP option 60, and all the options marked with the # vendor class are send encapsulated in DHCP option 43. The meaning of # the options is defined by the vendor-class. This example sets the # mtftp address to 0.0.0.0 for PXEClients #dhcp-option=vendor:PXEClient,1,0.0.0.0 # Set the boot filename and tftpd server name and address # for BOOTP. You will only need this is you want to # boot machines over the network. #dhcp-boot=/var/ftpd/pxelinux.0,boothost,192.168.0.3 # Set the limit on DHCP leases, the default is 150 #dhcp-lease-max=150 # The DHCP server needs somewhere on disk to keep its lease database. # This defaults to a sane location, but if you want to change it, use # the line below. #dhcp-leasefile=/var/lib/misc/dnsmasq.leases # Set the DHCP server to authoritative mode. In this mode it will barge in # and take over the lease for any client which broadcasts on the network, # whether it has a record of the lease or not. This avoids long timeouts # when a machine wakes up on a new network. DO NOT enable this if there's # the slighest chance that you might end up accidentally configuring a DHCP # server for your campus/company accidentally. The ISC server uses the same # the same option, and this URL provides more information: # <a href="http://www.isc.org/index.pl?/sw/dhcp/authoritative.php" rel="nofollow">http://www.isc.org/index.pl?/sw/dhcp/authoritative.php</a> #dhcp-authoritative # Set the cachesize here. #cache-size=150 # If you want to disable negative caching, uncomment this. #no-negcache # Normally responses which come form /etc/hosts and the DHCP lease # file have Time-To-Live set as zero, which conventionally means # do not cache further. If you are happy to trade lower load on the # server for potentially stale date, you can set a time-to-live (in # seconds) here. #local-ttl= # If you want dnsmasq to detect attempts by Verisign to send queries # to unregistered .com and .net hosts to its sitefinder service and # have dnsmasq instead return the correct NXDOMAIN response, uncomment # this line. You can add similar lines to do the same for other # registries which have implemented wildcard A records. #bogus-nxdomain=64.94.110.11 # If you want to fix up DNS results from upstream servers, use the # alias option. This only works for IPv4. # This alias makes a result of 1.2.3.4 appear as 5.6.7.8 #alias=1.2.3.4,5.6.7.8 # and this maps 1.2.3.x to 5.6.7.x #alias=1.2.3.0,5.6.7.0,255.255.255.0 # Change these lines if you want dnsmasq to serve MX records. # Return an MX record named "maildomain.com" with target # servermachine.com and preference 50 #mx-host=maildomain.com,servermachine.com,50 # Set the default target for MX records created using the localmx option. #mx-target=servermachine.com # Return an MX record pointing to the mx-target for all local # machines. #localmx # Return an MX record pointing to itself for all local machines. #selfmx # Change the following lines if you want dnsmasq to serve SRV # records. These are useful if you want to serve ldap requests for # Active Directory and other windows-originated DNS requests. # See <a href="http://tools.ietf.org/html/rfc2782">RFC 2782</a>. # You may add multiple srv-host lines. # The fields are <name>,<target>,<port>,<priority>,<weight> # If the domain part if missing from the name (so that is just has the # service and protocol sections) then the domain given by the domain= # config option is used. (Note that expand-hosts does not need to be # set for this to work.) # A SRV record sending LDAP for the example.com domain to # ldapserver.example.com port 289 #srv-host=_ldap._tcp.example.com,ldapserver.example.com,389 # A SRV record sending LDAP for the example.com domain to # ldapserver.example.com port 289 (using domain=) #domain=example.com #srv-host=_ldap._tcp,ldapserver.example.com,389 # Two SRV records for LDAP, each with different priorities #srv-host=_ldap._tcp.example.com,ldapserver.example.com,389,1 #srv-host=_ldap._tcp.example.com,ldapserver.example.com,389,2 # A SRV record indicating that there is no LDAP server for the domain # example.com #srv-host=_ldap._tcp.example.com # Change the following lines to enable dnsmasq to serve TXT records. # These are used for things like SPF and zeroconf. (Note that the # domain-name expansion done for SRV records _does_not # occur for TXT records.) #Example SPF. #txt-record=example.com,v=spf1 a -all #Example zeroconf #txt-record=_http._tcp.example.com,name=value,paper=A4 # For debugging purposes, log each DNS query as it passes through # dnsmasq. #log-queries # Include a another lot of configuration options. #conf-file=/etc/dnsmasq.more.conf |
Memcached 1.4.5里面通过sasl提供了认证支持,编译的时候可以启用该功能,编译参数:
[memcached-1.4.5]# ./configure –help| grep -i sasl
–enable-sasl Enable SASL authentication
–enable-sasl-pwdb Enable plaintext password db
SASL全称Simple Authentication and Security Layer,是一种用来扩充C/S模式验证能力的机制;
关于sasl:
参考文档:http://tools.ietf.org/html/rfc4422
c库实现: http://asg.web.cmu.edu/sasl/sasl-library.html
GNU libgsasl实现: http://www.gnu.org/software/gsasl/
其它资料:
http://asg.web.cmu.edu/sasl/
memcached 1.2.8
-p <num> TCP port number to listen on (default: 11211)
-U <num> UDP port number to listen on (default: 11211, 0 is off)
-s <file> unix socket path to listen on (disables network support)
-a <mask> access mask for unix socket, in octal (default 0700)
-l <ip_addr> interface to listen on, default is INDRR_ANY
-d run as a daemon
-r maximize core file limit
-u <username> assume identity of <username> (only when run as root)
-m <num> max memory to use for items in megabytes, default is 64 MB
-M return error on memory exhausted (rather than removing items)
-c <num> max simultaneous connections, default is 1024
-k lock down all paged memory. Note that there is a
limit on how much memory you may lock. Trying to
allocate more than that would fail, so be sure you
set the limit correctly for the user you started
the daemon with (not for -u <username> user;
under sh this is done with ‘ulimit -S -l NUM_KB’).
-v verbose (print errors/warnings while in event loop)
-vv very verbose (also print client commands/reponses)
-h print this help and exit
-i print memcached and libevent license
-P <file> save PID in <file>, only used with -d option
-f <factor> chunk size growth factor, default 1.25
-n <bytes> minimum space allocated for key+value+flags, default 48
-R Maximum number of requests per event
limits the number of requests process for a given con nection
to prevent starvation. default 20
-b Set the backlog queue limit (default 1024)
——————————————————————-
memcached 1.4.5
-p <num> TCP port number to listen on (default: 11211)
-U <num> UDP port number to listen on (default: 11211, 0 is off)
-s <file> UNIX socket path to listen on (disables network support)
-a <mask> access mask for UNIX socket, in octal (default: 0700)
-l <ip_addr> interface to listen on (default: INADDR_ANY, all addresses)
-d run as a daemon
-r maximize core file limit
-u <username> assume identity of <username> (only when run as root)
-m <num> max memory to use for items in megabytes (default: 64 MB)
-M return error on memory exhausted (rather than removing items)
-c <num> max simultaneous connections (default: 1024)
-k lock down all paged memory. Note that there is a
limit on how much memory you may lock. Trying to
allocate more than that would fail, so be sure you
set the limit correctly for the user you started
the daemon with (not for -u <username> user;
under sh this is done with ‘ulimit -S -l NUM_KB’).
-v verbose (print errors/warnings while in event loop)
-vv very verbose (also print client commands/reponses)
-vvv extremely verbose (also print internal state transitions)
-h print this help and exit
-i print memcached and libevent license
-P <file> save PID in <file>, only used with -d option
-f <factor> chunk size growth factor (default: 1.25)
-n <bytes> minimum space allocated for key+value+flags (default: 48)
-L Try to use large memory pages (if available). Increasing
the memory page size could reduce the number of TLB misses
and improve the performance. In order to get large pages
from the OS, memcached will allocate the total item-cache
in one large chunk.
-D <char> Use <char> as the delimiter between key prefixes and IDs.
This is used for per-prefix stats reporting. The default is
":" (colon). If this option is specified, stats collection
is turned on automatically; if not, then it may be turned on
by sending the "stats detail on" command to the server.
-t <num> number of threads to use (default: 4)
-R Maximum number of requests per event, limits the number of
requests process for a given connection to prevent
starvation (default: 20)
-C Disable use of CAS
-b Set the backlog queue limit (default: 1024)
-B Binding protocol – one of ascii, binary, or auto (default)
-I Override the size of each slab page. Adjusts max item size
(default: 1mb, min: 1k, max: 128m)
对于新的版本,增加了多线程处理机制,这样设定可能会提高memcache官方提供95%的命中率的问题。
对于memcache的使用,尽量使用内网,减少服务器连接的时间。
对于高并发的处理,memcache如果效果不太理想,可以尝试使用magent,memcache负载均衡。
使用PHP的memcache模块写了一个访问tokyotrant的long-live程序,因为是long-live的,所以我就connect一次之后一直使用了,理论上我connect之后就可以一直使用,中间不会出现重新连接的问题,为了确认我的推断,启动进程之后,我用strace跟踪了一些进程,令我意外的是,隔一段时间连接就会关闭,然后重新连接,怎么回事呢?
我怀疑两个方面:
1. 我的程序有问题
2. server端有问题,用一段时间会关掉我的连接
首先,我用了大约1个小时的时间,简直把我的程序拆的支离破碎了,结果没有发现哪里有问题。
其次,我使用tcpdump观察了一下,发现主动关闭连接的不是server端,而是我的程序。
百思不得其解。这件事简直成了我的一块心病。
隔了一段时间,当我再次使用tcpdump观察的时候,结果如下:
1. 21:27:20.273522 IP 10.55.38.62.60953 > 10.55.38.70.2004: . ack 132 win 501 <nop,nop,timestamp 3026990738 3683027032>
2. 21:27:20.273757 IP 10.55.38.62.60953 > 10.55.38.70.2004: P 20:142(122) ack 132 win 501 <nop,nop,timestamp 3026990738 3683027032>
3. 21:27:20.273871 IP 10.55.38.70.2004 > 10.55.38.62.60953: . ack 142 win 255 <nop,nop,timestamp 3683027032 3026990738>
4. 21:27:21.273955 IP 10.55.38.62.60953 > 10.55.38.70.2004: F 142:142(0) ack 132 win 501 <nop,nop,timestamp 3026991738 3683027032>
5. 21:27:21.274038 IP 10.55.38.62.37298 > 10.55.38.70.2004: S 3948732568:3948732568(0) win 5792 <mss 1460,sackOK,timestamp 3026991738 3683027542,nop,wscale 7>
6. 21:27:21.274165 IP 10.55.38.70.2004 > 10.55.38.62.37298: S 34634952:34634952(0) ack 3948732569 win 5792 <mss 1460,sackOK,timestamp 3683028032 3026991738,nop,wscale 7>
21:27:21.274185 IP 10.55.38.62.37298 > 10.55.38.70.2004: . ack 1 win 46 <nop,nop,timestamp 3026991738 3683028032>
21:27:21.274196 IP 10.55.38.62.37298 > 10.55.38.70.2004: P 1:123(122) ack 1 win 46 <nop,nop,timestamp 3026991738 3683028032>
21:27:21.274320 IP 10.55.38.70.2004 > 10.55.38.62.37298: . ack 123 win 46 <nop,nop,timestamp 3683028032 3026991738>
21:27:21.313329 IP 10.55.38.70.2004 > 10.55.38.62.60953: . ack 143 win 255 <nop,nop,timestamp 3683028072 3026991738>
21:27:21.533806 IP 10.55.38.70.2004 > 10.55.38.62.37298: P 1:9(8) ack 123 win 46 <nop,nop,timestamp 3683028292 3026991738>
21:27:21.533822 IP 10.55.38.62.37298 > 10.55.38.70.2004: . ack 9 win 46 <nop,nop,timestamp 3026991998 3683028292>
21:27:22.604843 IP 10.55.38.70.2004 > 10.55.38.62.60953: P 132:140(8) ack 143 win 255 <nop,nop,timestamp 3683029363 3026991738>
21:27:22.604859 IP 10.55.38.62.60953 > 10.55.38.70.2004: R 3898017016:3898017016(0) win 0
21:27:24.072995 IP 10.55.38.62.37298 > 10.55.38.70.2004: P 123:143(20) ack 9 win 46 <nop,nop,timestamp 3026994537 3683028292>
============================================
我意外地发现有一个reset包,感觉很奇怪,顺着 60953 端口网上查,发现:
第2个包: 向server端发送数据
第3个包: server端回复收到数据
第4个包: 按说应该server端返回响应数据,但是这里却是client端发了一个finish包
观察第3个包与第4个包之间的时间间隔,基本是1s, 这令我想起了memcache的默认超时时间也是1s,是巧合?显然不是。因为:
第5个包: client端重新发起了连接操作
显然是超时了, 于是重连有了答案。
当然,虽然client关闭了连接,server端却还没有关闭,直到 21:27:22.604859 的时候,服务器端给出了响应数据,但是client已经关闭了,所以出现了被reset的现象。
==========================================
还有一个问题:
我的程序里面是做了容错处理的,但是我的容错策略是,如果出错了,则sleep(2);之后重新连接; 显然不符合上面的表现,怀着万般不解的心情开始看php的memcache模块的源码;
在memcache.c中:
在这里,我们发现,如果store失败的话,下面会有一个容错处理,通过php -i | grep memcache 可以看到容错配置确实是打开的。
在容错的代码中明显发现做了断开连接的操作,但是在哪里重连的呢?
这要看mmc_hash_find_server的实现了,有时间再看。。。