blob: 5076f0d78e9389a1a3fd022eb2003e3ca985a24b (
plain)
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
|
# See the file LICENSE for redistribution information.
#
# Copyright (c) 2003, 2012 Oracle and/or its affiliates. All rights reserved.
#
# $Id$
#
# TEST rep100
# TEST Checkpoints and unresolved txns
# TEST
proc rep100 { method { niter 10 } { tnum "100" } args } {
source ./include.tcl
global repfiles_in_memory
# Run for all access methods.
if { $checking_valid_methods } {
return "ALL"
}
set args [convert_args $method $args]
set logsets [create_logsets 2]
set msg2 "and on-disk replication files"
if { $repfiles_in_memory } {
set msg2 "and in-memory replication files"
}
# Run the body of the test with and without recovery.
foreach r $test_recopts {
foreach l $logsets {
set logindex [lsearch -exact $l "in-memory"]
if { $logindex != -1 } {
puts "Rep$tnum: Skipping for in-memory logs."
continue
}
puts "Rep$tnum ($method $r):\
Checkpoints and unresolved txns $msg2."
puts "Rep$tnum: Master logs are [lindex $l 0]"
puts "Rep$tnum: Client logs are [lindex $l 1]"
rep100_sub $method $niter $tnum $l $r $args
}
}
}
proc rep100_sub { method niter tnum logset recargs largs } {
source ./include.tcl
global perm_response_list
global repfiles_in_memory
global rep_verbose
global verbose_type
set verbargs ""
if { $rep_verbose == 1 } {
set verbargs " -verbose {$verbose_type on} "
}
set repmemargs ""
if { $repfiles_in_memory } {
set repmemargs "-rep_inmem_files "
}
env_cleanup $testdir
set omethod [convert_method $method]
replsetup $testdir/MSGQUEUEDIR
set masterdir $testdir/MASTERDIR
set clientdir $testdir/CLIENTDIR
file mkdir $masterdir
file mkdir $clientdir
set m_logtype [lindex $logset 0]
set c_logtype [lindex $logset 1]
# Since we're sure to be using on-disk logs, txnargs will be -txn nosync.
set m_logargs [adjust_logargs $m_logtype]
set c_logargs [adjust_logargs $c_logtype]
set m_txnargs [adjust_txnargs $m_logtype]
set c_txnargs [adjust_txnargs $c_logtype]
# Open a master.
repladd 1
set ma_cmd "berkdb_env_noerr -create $verbargs \
-log_max 1000000 $m_txnargs $m_logargs $repmemargs \
-home $masterdir -rep_master -errpfx MASTER \
-rep_transport \[list 1 replsend\]"
set masterenv [eval $ma_cmd $recargs]
# Open a client
repladd 2
set cl_cmd "berkdb_env_noerr -create -home $clientdir $verbargs \
$c_txnargs $c_logargs -rep_client -errpfx CLIENT $repmemargs \
-rep_transport \[list 2 replsend\]"
set clientenv [eval $cl_cmd $recargs]
# Bring the client online.
process_msgs "{$masterenv 1} {$clientenv 2}"
# Open database in master, make lots of changes so checkpoint
# will take a while, and propagate to client.
puts "\tRep$tnum.a: Create and populate database."
set dbname rep100.db
set dbname1 rep100_1.db
set db [eval "berkdb_open_noerr -create $omethod -auto_commit \
-env $masterenv $largs $dbname"]
set db1 [eval "berkdb_open_noerr -create $omethod -auto_commit \
-env $masterenv $largs $dbname1"]
for { set i 1 } { $i <= $niter } { incr i } {
set t [$masterenv txn]
error_check_good db_put \
[eval $db put -txn $t $i [chop_data $method data$i]] 0
error_check_good db1_put \
[eval $db1 put -txn $t $i [chop_data $method data$i]] 0
error_check_good txn_commit [$t commit] 0
}
process_msgs "{$masterenv 1} {$clientenv 2}" 1
# Get the master's last LSN before the checkpoint
set pre_ckp_offset \
[stat_field $masterenv log_stat "Current log file offset"]
puts "\tRep$tnum.b: Checkpoint on master."
error_check_good checkpoint [$masterenv txn_checkpoint] 0
process_msgs "{$masterenv 1} {$clientenv 2}"
#
# We want to generate a checkpoint that is mid-txn on the master,
# but is mid-txn on a different txn on the client because the
# client is behind the master. We want to make sure we don't
# get a Log sequence error on recovery. The sequence of steps is:
#
# Open a txn T1 on the master. Made a modification.
# Open a 2nd txn T2 on the master and make a modification for that txn.
# Replicate all of the above to the client.
# Make another modification to T1.
# Commit T1 but do not replicate to the client (i.e. lose those records).
# Checkpoint on the master and replicate to client.
# This should cause the client to sync pages but not the pages that
# modify T1 because its commit hasn't appeared yet, even though it
# has committed on the master.
# Commit T2
# Crash client and run recovery.
#
set start $niter
set end [expr $niter * 2]
# Open txn T1 and make a modification.
puts "\tRep$tnum.c: Open txn t1 and do an update."
set i $start
set t1 [$masterenv txn]
error_check_good db_put \
[eval $db1 put -txn $t1 $i [chop_data $method data$i]] 0
# Open a 2nd txn T2 on the master and make a modification in that txn.
# Replicate all to the client.
puts "\tRep$tnum.d: Open txn T2."
set t2 [$masterenv txn]
error_check_good db_put \
[eval $db put -txn $t2 $end [chop_data $method data$end]] 0
process_msgs "{$masterenv 1} {$clientenv 2}"
# Make another modification to T1.
# Commit T1 but do not replicate to the client (i.e. lose that record).
puts "\tRep$tnum.e: Update and commit T1, clear msgs for client."
error_check_good db_put \
[eval $db1 put -txn $t1 $i [chop_data $method data$end]] 0
error_check_good txn_commit [$t1 commit] 0
replclear 2
# Checkpoint on the master and replicate to client.
puts "\tRep$tnum.f: Checkpoint and replicate messages."
error_check_good checkpoint [$masterenv txn_checkpoint] 0
process_msgs "{$masterenv 1} {$clientenv 2}"
#
# Commit T2. We have to sleep enough time so that the client
# will rerequest and catch up when it receives these records.
#
puts "\tRep$tnum.g: Sleep, commit T2 and catch client up."
tclsleep 2
error_check_good txn_commit [$t2 commit] 0
process_msgs "{$masterenv 1} {$clientenv 2}"
#
# At this point the client has all outstanding log records that
# have been written by the master. So everything should be okay.
#
puts "\tRep$tnum.h: Abandon clientenv and reopen with recovery."
# "Crash" client (by abandoning its env) and run recovery.
set clientenv_new [eval $cl_cmd -recover]
# Clean up.
puts "\tRep$tnum.i: Clean up."
error_check_good db_close [$db close] 0
error_check_good db_close [$db1 close] 0
error_check_good masterenv_close [$masterenv close] 0
error_check_good clientenv_close [$clientenv_new close] 0
# Clean up abandoned handle.
catch {$clientenv close}
replclose $testdir/MSGQUEUEDIR
}
|
