註冊 用戶 超過 130000 178864
設備 156584
1.
測試關於“漢字”更新數據
後臺拋出 =INFO REPORT==== 2010-10-25 18:07:47 ===
D(<0.1906.0>:ejabberd_odbc:416) : MySQL, Send query
["UPDATE vendors SET vendorname='",[20013],"' WHERE vendorid='","37","';"]
=INFO REPORT==== 2010-10-25 18:07:52 ===
D(<0.1906.0>:ejabberd_odbc:416) : MySQL, Send query
"rollback;"
解決方法：
[20013] 是爲調用了xmerl:xmerl_scan:string/2 返回的,只好從編碼上入手，但
很不能理解 如果 下面參數 S 已經是UTF8 了在執行下面操作後 又進行了編碼（爲什麼再次編碼，此現象只出現在漢字上面）。所以只好在針對xmerl_scan:string/2的返回值進行編碼，而不能寫成通用的方法。
如[233,223,222] 執行下面操作後成了 ［228,123,345,238,222,133］(此數據是說明問題隨便寫的)
> > + UTF8Binary = unicode:characters_to_binary(S, utf8, utf8),
> > + UTF8Str = binary_to_list(UTF8Binary),
2.單個 虛擬機 測試（ <client host="client-3.langzhelabdns.com" use_controller_vm="true" maxusers="60000"/>
）
解決方法：以下問題 修改ulimit -n 102400參數解決
INFO REPORT==== 1-Nov-2010::13:46:29 ===
ts_launcher:(5:<0.84.0>) Activate launcher (60000 users) in 10002 msec
=INFO REPORT==== 1-Nov-2010::13:49:47 ===
ts_client:(3:<0.12537.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: etimedout
=INFO REPORT==== 1-Nov-2010::13:49:47 ===
ts_client:(3:<0.12536.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: etimedout
=INFO REPORT==== 1-Nov-2010::13:49:47 ===
ts_client:(3:<0.12535.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: etimedout
日誌－－－：
ts_client:(3:<0.6577.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: system_limit
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
ts_client:(3:<0.4203.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: system_limit
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
ts_client:(5:<0.4203.0>) Pending transactions: [{tr_device_register,
{1288,578941,
906458}}], compute transaction time
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
ts_client:(3:<0.4378.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: system_limit
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
ts_client:(5:<0.4378.0>) Pending transactions: [{tr_device_register,
{1288,578941,
906438}}], compute transaction time
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
ts_client:(3:<0.5622.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: system_limit
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
ts_client:(3:<0.4033.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: system_limit
=INFO REPORT==== 1-Nov-2010::10:35:44 ===
Error: econnreset
ts_client:(3:<0.32258.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: econnrefused
=INFO REPORT==== 1-Nov-2010::10:41:03 ===
ts_client:(3:<0.27051.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: econnrefused
=INFO REPORT==== 1-Nov-2010::10:41:03 ===
ts_client:(5:<0.27051.0>) Pending transactions: [{tr_user_register,
{1288,579203,
9045}}], compute transaction time
=INFO REPORT==== 1-Nov-2010::10:41:03 ===
ts_client:(3:<0.31393.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: econnrefused
3. ------------------------------------------------
根據上面出現的情況 調整ulimit參數後，tsung日誌出現 etimedout （ssh驗證 tsung監聽 ）會出現此問題
參見 http://www.iteye.com/topic/573238部分解釋
=INFO REPORT==== 1-Nov-2010::13:49:47 ===
ts_client:(3:<0.12532.0>) (Re)connect from 0.0.0.0 to testrsa.langzhelabdns.com:5280, Error: etimedout
解決方法 配置ssh 信任
4. －－－－死鎖 ------------------ -------------代碼中出現了for update引起 去掉就OK了-- ---------------
=INFO REPORT==== 2010-11-01 10:22:06 ===
15403 D(<0.517.0>:ejabberd_http:323) : ["user","register"] matches ["user",
15404 "register"]
15405
15406 =INFO REPORT==== 2010-11-01 10:22:06 ===
15407 D(<0.259.0>:ejabberd_odbc:420) : MySQL, Send query
15408 ["SELECT vendorid FROM vendors WHERE vendorname = '","test","' FOR UPDATE;"]
15409
15410
15411 =INFO REPORT==== 2010-11-01 10:22:06 ===
15412 D(<0.247.0>:ejabberd_odbc:420) : MySQL, Send query
15413 ["SELECT vendorid FROM vendors WHERE vendorname = '","test","' FOR UPDATE;"]
15414
15415
15416 =INFO REPORT==== 2010-11-01 10:22:06 ===
15417 D(<0.437.0>:mod_langzhe_device_register:71) : transactionResult = {aborted,
15418 {{badmatch,
15419 {error,
15420 "#40001Deadlock found when trying to get lock; try restarting transaction"}},
15421 [{mod_langzhe_basic,
15422 add_device_type,
15423 2},
15424 {mod_langzhe_device_register,
15425 '-process/2-fun-0-',
15426 7},
15427 {ejabberd_odbc,
15428 outer_transaction,
15429 3},
15430 {ejabberd_odbc,
15431 session_established,
15432 3},
15433 {p1_fsm,
15434 handle_msg,
15435 10},
5.－－－－－－－－－－－ 註冊用戶 40080－ －－－－－－－－－－－－－
Mon Nov 1 16:05:19 2010
Slogan: eheap_alloc: Cannot allocate 729810240 bytes of memory (of type "heap").
System version: Erlang R13B03 (erts-5.7.4) [source] [smp:2:2] [rq:2] [async-threads:0] [kernel-poll:true]
Compiled: Fri Oct 15 14:49:18 2010
Atoms: 15554
=memory
total: 1802636424
processes: 1646297908
processes_used: 1645947788
system: 156338516
atom: 710173
atom_used: 696317
binary: 235232
code: 6574357
ets: 5148228
=hash_table:atom_tab
＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝
Mon Nov 1 15:52:53 2010
Slogan: eheap_alloc: Cannot allocate 912262800 bytes of memory (of type "heap").
System version: Erlang R13B03 (erts-5.7.4) [source] [smp:2:2] [rq:2] [async-threads:0] [kernel-poll:true]
Compiled: Fri Oct 15 14:49:18 2010
Atoms: 15554
=memory
total: 2105409384
processes: 2059069028
processes_used: 2058770452
system: 46340356
atom: 710173
atom_used: 696317
binary: 207192
code: 6574357
ets: 5148596
=hash_table:atom_tab
size: 9643
used: 7694
objs: 15554
depth: 8
-----------------------------------------
=erl_crash_dump:0.1註冊用戶數 32874 600s 300/s 服務器拋出 錯誤時 物理內存 被佔用81％ （2G內存）
912262800/1024/1024 > 0.19*2*1024(870>389 顯然 是內存不足)
Mon Nov 1 16:30:25 2010
Slogan: eheap_alloc: Cannot allocate 912262800 bytes of memory (of type "heap").
System version: Erlang R13B03 (erts-5.7.4) [source] [smp:2:2] [rq:2] [async-threads:0] [kernel-poll:true]
Compiled: Fri Oct 15 14:49:18 2010
Atoms: 15554
=memory
total: 2238866320
processes: 2077803940
processes_used: 2077676852
system: 161062380
atom: 710173
atom_used: 696317
binary: 222712
code: 6574357
ets: 5148340
=hash_table:atom_tab
size: 9643
used: 7694
objs: 15554
depth: 8
解決方法
分析後發現是日誌級別引起，設爲info就OK
六、－－－－－－－－－－－FSM limit reached (<0.255.0>): {max_queue,1003}--------------------------------------------
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.27300.0>:ejabberd_http:137) : started: {gen_tcp,#Port<0.30873>}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.384.0>:ejabberd_listener:234) : (#Port<0.30874>) Accepted connection {{192,168,0,223},40244} -> {{192,168,0,223},5280}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.27301.0>:ejabberd_http:137) : started: {gen_tcp,#Port<0.30874>}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.384.0>:ejabberd_listener:234) : (#Port<0.30875>) Accepted connection {{192,168,0,223},54723} -> {{192,168,0,223},5280}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.27302.0>:ejabberd_http:137) : started: {gen_tcp,#Port<0.30875>}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.384.0>:ejabberd_listener:234) : (#Port<0.30876>) Accepted connection {{192,168,0,223},47419} -> {{192,168,0,223},5280}
=ERROR REPORT==== 2010-11-01 20:30:30 ===
FSM limit reached (<0.255.0>): {max_queue,1003}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.257.0>:ejabberd_odbc:526) : Closing connection #Port<0.3650>: ok
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.27303.0>:ejabberd_http:137) : started: {gen_tcp,#Port<0.30876>}
=INFO REPORT==== 2010-11-01 20:30:30 ===
I(<0.384.0>:ejabberd_listener:234) : (#Port<0.30877>) Accepted connection {{192,168,0,223},60273} -> {{192,168,0,223},5280}
=INFO REPORT==== 2010-11-01 20:30:33 ===
I(<0.27307.0>:ejabberd_http:137) : started: {gen_tcp,#Port<0.30877>}
=INFO REPORT==== 2010-11-01 20:30:34 ===
I(<0.384.0>:ejabberd_listener:234) : (#Port<0.30878>) Accepted connectio
解決方法 參見 : http://langzhe.iteye.com/admin/blogs/801942
七、－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－
=INFO REPORT==== 2010-11-02 08:57:24 ===
I(<0.384.0>:ejabberd_listener:246) : (#Port<0.4265>) Failed TCP accept: emfile
解決方法: vi /etc/ejabberd/ejabberdctl.cfg 修改
ERL_PROCESSES=250000
ERL_MAX_PORTS=320000
參數和ulimit -n 參數
enfile問題參考 http://www.netadmintools.com/art295.html
八、－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－
方法1 修改maxusers參數<client host="client-3.langzhelabdns.com" use_controller_vm="true" maxusers="6000000"/>
方法2 修改use_controller_vm參數 <client host="client-3.langzhelabdns.com" use_controller_vm="faluse"/>
"Maximum number of concurrent users in a single VM reached and 'use_controller_vm' is true, can't start new beam !!! Check 'maxusers' value in <client> configuration.~n"
九、－－－－－－－－－－－ID 越過－－－沒有生成 vendordid 爲64的記錄－－－－－－－－－－－－－－－－－－－－－－
select * from vendors where vendorid like'6_' order by vendorid;
+----------+------------+
| vendorid | vendorname |
+----------+------------+
| 60 | test160 |
| 61 | test161 |
| 62 | test162 |
| 63 | test163 |
| 65 | test164 |
| 66 | test165 |
| 67 | test166 |
| 68 | test167 |
| 69 | test168 |
+----------+------------+
157 | test1157 |
| 159 | test1158 |
| 160 | test1159
＝＝＝＝＝＝下面兩個問題還沒有解決＝＝＝＝＝＝＝＝＝＝＝＝
十一、－－－－－－－－－－－－－－－－－－－－－－－－－－
=INFO REPORT==== 2010-11-03 17:42:19 ===
I(<0.20211.0>:ejabberd_http:137) : started: {gen_tcp,#Port<0.23590>}
=ERROR REPORT==== 2010-11-03 17:42:19 ===
** State machine <0.235.0> terminating
** Last message in was {'$gen_sync_event',
{<0.7013.0>,#Ref<0.0.0.125463>},
{sql_cmd,
{sql_query,
["SELECT keyt FROM langzhe_vendor_key WHERE vendor = '",
"test","' and version = '","1","';"]}}}
** When State == session_established
** Data == {state,<0.236.0>,mysql,30000,"testrsa.langzhelabdns.com",
1000000000000,
{0,{[],[]}}}
** Reason for termination =
** timeout
==> ejabberd.log <==
{sql_cmd,
{sql_query,
["SELECT keyt FROM langzhe_vendor_key WHERE vendor = '",
"test","' and version = '","1","';"]}}}
** When State == session_established
** Data == {state,<0.236.0>,mysql,30000,"testrsa.langzhelabdns.com",
1000000000000,
{0,{[],[]}}}
** Reason for termination =
** timeout
十二、－－－－－－－－－－－－－－－－－－－－－mysql connection failed timeout－－－－－－－－－－－－－－－
5 =INFO REPORT==== 2010-11-05 16:23:21 ===
6 I(<0.247.0>:ejabberd_odbc:212) : mysql connection failed:
7 ** Reason: "timed out"
8 ** Retry after: 30 seconds
9
10 =ERROR REPORT==== 2010-11-05 16:23:21 ===
11 E(<0.246.0>:ejabberd_odbc:528) : mysql_conn: Failed connecting to "192.168.0.223":3306 : {error,
12 "timeout"}
13
14 =INFO REPORT==== 2010-11-05 16:23:21 ===
15 I(<0.244.0>:ejabberd_odbc:212) : mysql connection failed:
16 ** Reason: "timed out"
17 ** Retry after: 30 seconds
18
19 =ERROR REPORT==== 2010-11-05 16:23:21 ===
20 W(<0.244.0>:ejabberd_odbc:281) : unexpected info in connecting: {mysql_conn,
21 <0.246.0>,
22 {error,
23 connect_failed}}
24
25 =ERROR REPORT==== 2010-11-05 16:23:21 ===
26 E(<0.243.0>:ejabberd_odbc:528) : mysql_conn: Failed connecting to "192.168.0.223":3306 : {error,
27 "timeout"}
28
29 =INFO REPORT==== 2010-11-05 16:23:21 ===
30 I(<0.241.0>:ejabberd_odbc:212) : mysql connection failed:
31 ** Reason: "timed out"
32 ** Retry after: 30 seconds
33
34 =ERROR REPORT==== 2010-11-05 16:23:21 ===
35 W(<0.241.0>:ejabberd_odbc:281) : unexpected info in connecting: {mysql_conn,
36 <0.243.0>,
37 {error,
