| 1 |
# SOME DESCRIPTIVE TITLE |
| 2 |
# Copyright (C) YEAR Free Software Foundation, Inc. |
| 3 |
# This file is distributed under the same license as the PACKAGE package. |
| 4 |
# FIRST AUTHOR <EMAIL@ADDRESS>, YEAR. |
| 5 |
# |
| 6 |
#, fuzzy |
| 7 |
msgid "" |
| 8 |
msgstr "" |
| 9 |
"Project-Id-Version: PACKAGE VERSION\n" |
| 10 |
"POT-Creation-Date: 2020-03-02 15:09+0000\n" |
| 11 |
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n" |
| 12 |
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n" |
| 13 |
"Language-Team: LANGUAGE <[email protected]>\n" |
| 14 |
"Language: \n" |
| 15 |
"MIME-Version: 1.0\n" |
| 16 |
"Content-Type: text/plain; charset=UTF-8\n" |
| 17 |
"Content-Transfer-Encoding: 8bit\n" |
| 18 |
|
| 19 |
#. type: Title == |
| 20 |
#: benchmark.adoc:2 |
| 21 |
#, no-wrap |
| 22 |
msgid "Measuring block device performance" |
| 23 |
msgstr "测量块设备性能" |
| 24 |
|
| 25 |
#. type: Title === |
| 26 |
#: benchmark.adoc:6 |
| 27 |
#, no-wrap |
| 28 |
msgid "Measuring throughput" |
| 29 |
msgstr "测量吞吐量" |
| 30 |
|
| 31 |
#. type: Plain text |
| 32 |
#: benchmark.adoc:13 |
| 33 |
msgid "" |
| 34 |
"When measuring the impact of using DRBD on a system's I/O throughput, the " |
| 35 |
"_absolute_ throughput the system is capable of is of little relevance. What " |
| 36 |
"is much more interesting is the _relative_ impact DRBD has on I/O " |
| 37 |
"performance. Thus it is always necessary to measure I/O throughput both with" |
| 38 |
" and without DRBD." |
| 39 |
msgstr "" |
| 40 |
"当测量使用DRBD对系统I/O吞吐量的影响时,系统所能达到的绝对吞吐量几乎没有相关性。更有趣的是DRBD对I/O性能的相对影响。因此,无论有无DRBD,总是有必要测量I/O吞吐量。" |
| 41 |
|
| 42 |
#. type: Plain text |
| 43 |
#: benchmark.adoc:18 |
| 44 |
msgid "" |
| 45 |
"The tests described in this section are intrusive; they overwrite data and " |
| 46 |
"bring DRBD devices out of sync. It is thus vital that you perform them only " |
| 47 |
"on scratch volumes which can be discarded after testing has completed." |
| 48 |
msgstr "本节中描述的测试是侵入性的;它们覆盖数据并使DRBD设备失去同步。因此,只在测试完成后可以丢弃的暂存卷上执行它们是非常重要的。" |
| 49 |
|
| 50 |
#. type: Plain text |
| 51 |
#: benchmark.adoc:24 |
| 52 |
msgid "" |
| 53 |
"I/O throughput estimation works by writing significantly large chunks of " |
| 54 |
"data to a block device, and measuring the amount of time the system took to " |
| 55 |
"complete the write operation. This can be easily done using a fairly " |
| 56 |
"ubiquitous utility, `dd`, whose reasonably recent versions include a built-" |
| 57 |
"in throughput estimation." |
| 58 |
msgstr "" |
| 59 |
"I/O吞吐量估计的工作原理是将大量数据写入块设备,并测量系统完成写入操作所需的时间。这可以使用相当普遍的实用程序 `dd` 轻松完成,该实用程序的最新版本包括内置的吞吐量估计。" |
| 60 |
|
| 61 |
#. type: Plain text |
| 62 |
#: benchmark.adoc:28 |
| 63 |
msgid "" |
| 64 |
"A simple ``dd``-based throughput benchmark, assuming you have a scratch " |
| 65 |
"resource named `test`, which is currently connected and in the secondary " |
| 66 |
"role on both nodes, is one like the following:" |
| 67 |
msgstr "一个简单的基于 `dd` 的吞吐量基准,假设您有一个名为 `test` 的临时资源,该资源当前已连接并且在两个节点上都处于次要角色,如下所示:" |
| 68 |
|
| 69 |
#. type: delimited block - |
| 70 |
#: benchmark.adoc:42 |
| 71 |
#, no-wrap |
| 72 |
msgid "" |
| 73 |
"# TEST_RESOURCE=test\n" |
| 74 |
"# TEST_DEVICE=$(drbdadm sh-dev $TEST_RESOURCE | head -1)\n" |
| 75 |
"# TEST_LL_DEVICE=$(drbdadm sh-ll-dev $TEST_RESOURCE | head -1)\n" |
| 76 |
"# drbdadm primary $TEST_RESOURCE\n" |
| 77 |
"# for i in $(seq 5); do\n" |
| 78 |
" dd if=/dev/zero of=$TEST_DEVICE bs=1M count=512 oflag=direct\n" |
| 79 |
" done\n" |
| 80 |
"# drbdadm down $TEST_RESOURCE\n" |
| 81 |
"# for i in $(seq 5); do\n" |
| 82 |
" dd if=/dev/zero of=$TEST_LL_DEVICE bs=1M count=512 oflag=direct\n" |
| 83 |
" done\n" |
| 84 |
msgstr "" |
| 85 |
"# TEST_RESOURCE=test\n" |
| 86 |
"# TEST_DEVICE=$(drbdadm sh-dev $TEST_RESOURCE | head -1)\n" |
| 87 |
"# TEST_LL_DEVICE=$(drbdadm sh-ll-dev $TEST_RESOURCE | head -1)\n" |
| 88 |
"# drbdadm primary $TEST_RESOURCE\n" |
| 89 |
"# for i in $(seq 5); do\n" |
| 90 |
" dd if=/dev/zero of=$TEST_DEVICE bs=1M count=512 oflag=direct\n" |
| 91 |
" done\n" |
| 92 |
"# drbdadm down $TEST_RESOURCE\n" |
| 93 |
"# for i in $(seq 5); do\n" |
| 94 |
" dd if=/dev/zero of=$TEST_LL_DEVICE bs=1M count=512 oflag=direct\n" |
| 95 |
" done\n" |
| 96 |
|
| 97 |
#. type: Plain text |
| 98 |
#: benchmark.adoc:48 |
| 99 |
msgid "" |
| 100 |
"This test simply writes 512MiB of data to your DRBD device, and then to its " |
| 101 |
"backing device for comparison. Both tests are repeated 5 times each to allow" |
| 102 |
" for some statistical averaging. The relevant result is the throughput " |
| 103 |
"measurements generated by `dd`." |
| 104 |
msgstr "" |
| 105 |
"此测试只需将512MiB的数据写入DRBD设备,然后写入其备份设备进行比较。这两项测试各重复5次,以便进行一些统计平均。相关结果是由 `dd` 生成的吞吐量测量。" |
| 106 |
|
| 107 |
#. type: Plain text |
| 108 |
#: benchmark.adoc:52 |
| 109 |
msgid "" |
| 110 |
"For freshly enabled DRBD devices, it is normal to see slightly reduced " |
| 111 |
"performance on the first `dd` run. This is due to the Activity Log being " |
| 112 |
"\"cold\", and is no cause for concern." |
| 113 |
msgstr "对于刚刚启用的DRBD设备,在第一次 `dd` 运行时性能会略有下降是正常的。这是因为活动日志是 `冷` 的,不需要担心。" |
| 114 |
|
| 115 |
#. type: Plain text |
| 116 |
#: benchmark.adoc:54 |
| 117 |
msgid "See our <<ch-throughput>> chapter for some performance numbers." |
| 118 |
msgstr "有关一些性能数据,请参阅我们的 <<ch-throughput>> 一章。" |
| 119 |
|
| 120 |
#. type: Title === |
| 121 |
#: benchmark.adoc:57 |
| 122 |
#, no-wrap |
| 123 |
msgid "Measuring latency" |
| 124 |
msgstr "测量延迟" |
| 125 |
|
| 126 |
#. type: Plain text |
| 127 |
#: benchmark.adoc:65 |
| 128 |
msgid "" |
| 129 |
"Latency measurements have objectives completely different from throughput " |
| 130 |
"benchmarks: in I/O latency tests, one writes a very small chunk of data " |
| 131 |
"(ideally the smallest chunk of data that the system can deal with), and " |
| 132 |
"observes the time it takes to complete that write. The process is usually " |
| 133 |
"repeated several times to account for normal statistical fluctuations." |
| 134 |
msgstr "" |
| 135 |
"延迟测量的目标与吞吐量基准完全不同:在I/O延迟测试中,一个人会写入非常小的数据块(理想情况下是系统可以处理的最小数据块),并观察完成该写入所需的时间。这个过程通常重复几次,以解释正常的统计波动。" |
| 136 |
|
| 137 |
#. type: Plain text |
| 138 |
#: benchmark.adoc:69 |
| 139 |
msgid "" |
| 140 |
"Just as throughput measurements, I/O latency measurements may be performed " |
| 141 |
"using the ubiquitous `dd` utility, albeit with different settings and an " |
| 142 |
"entirely different focus of observation." |
| 143 |
msgstr "与吞吐量测量一样,可以使用无处不在的 `dd` 实用程序执行I/O延迟测量,尽管设置不同,观察的焦点完全不同。" |
| 144 |
|
| 145 |
#. type: Plain text |
| 146 |
#: benchmark.adoc:73 |
| 147 |
msgid "" |
| 148 |
"Provided below is a simple ``dd``-based latency micro-benchmark, assuming " |
| 149 |
"you have a scratch resource named `test` which is currently connected and in" |
| 150 |
" the secondary role on both nodes:" |
| 151 |
msgstr "" |
| 152 |
"下面提供了一个简单的基于 `dd` 的延迟微基准测试,假设您有一个名为 `test` 的临时资源,该资源当前已连接,并且在两个节点上都处于次要角色:" |
| 153 |
|
| 154 |
#. type: delimited block - |
| 155 |
#: benchmark.adoc:83 |
| 156 |
#, no-wrap |
| 157 |
msgid "" |
| 158 |
"# TEST_RESOURCE=test\n" |
| 159 |
"# TEST_DEVICE=$(drbdadm sh-dev $TEST_RESOURCE | head -1)\n" |
| 160 |
"# TEST_LL_DEVICE=$(drbdadm sh-ll-dev $TEST_RESOURCE | head -1)\n" |
| 161 |
"# drbdadm primary $TEST_RESOURCE\n" |
| 162 |
"# dd if=/dev/zero of=$TEST_DEVICE bs=4k count=1000 oflag=direct\n" |
| 163 |
"# drbdadm down $TEST_RESOURCE\n" |
| 164 |
"# dd if=/dev/zero of=$TEST_LL_DEVICE bs=4k count=1000 oflag=direct\n" |
| 165 |
msgstr "" |
| 166 |
"# TEST_RESOURCE=test\n" |
| 167 |
"# TEST_DEVICE=$(drbdadm sh-dev $TEST_RESOURCE | head -1)\n" |
| 168 |
"# TEST_LL_DEVICE=$(drbdadm sh-ll-dev $TEST_RESOURCE | head -1)\n" |
| 169 |
"# drbdadm primary $TEST_RESOURCE\n" |
| 170 |
"# dd if=/dev/zero of=$TEST_DEVICE bs=4k count=1000 oflag=direct\n" |
| 171 |
"# drbdadm down $TEST_RESOURCE\n" |
| 172 |
"# dd if=/dev/zero of=$TEST_LL_DEVICE bs=4k count=1000 oflag=direct\n" |
| 173 |
|
| 174 |
#. type: Plain text |
| 175 |
#: benchmark.adoc:89 |
| 176 |
msgid "" |
| 177 |
"This test writes 1,000 chunks with 4kiB each to your DRBD device, and then " |
| 178 |
"to its backing device for comparison. 4096 bytes is the smallest block size " |
| 179 |
"that a Linux system (on all architectures except s390), modern hard disks, " |
| 180 |
"and SSDs, are expected to handle." |
| 181 |
msgstr "" |
| 182 |
"这个测试用4kiB分别将1000个数据块写入DRBD设备,然后写入其备份设备进行比较。4096字节是Linux系统(除了s390以外的所有体系结构)、现代硬盘和固态硬盘预计要处理的最小块大小。" |
| 183 |
|
| 184 |
#. type: Plain text |
| 185 |
#: benchmark.adoc:94 |
| 186 |
msgid "" |
| 187 |
"It is important to understand that throughput measurements generated by `dd`" |
| 188 |
" are completely irrelevant for this test; what is important is the _time_ " |
| 189 |
"elapsed during the completion of said 1,000 writes. Dividing this time by " |
| 190 |
"1,000 gives the average latency of a single block write." |
| 191 |
msgstr "" |
| 192 |
"重要的是要了解由 `dd` 生成的吞吐量测量与此测试完全无关;重要的是在完成所述1000次写入过程中经过的时间。将此时间除以1000可得出单个块写入的平均延迟。" |
| 193 |
|
| 194 |
#. type: Plain text |
| 195 |
#: benchmark.adoc:98 |
| 196 |
msgid "" |
| 197 |
"This is the _worst-case_, in that it is single-threaded and does one write " |
| 198 |
"strictly after the one before, i.e. runs with an I/O-depth of 1. Please take " |
| 199 |
"a look at <<s-latency-iops>>." |
| 200 |
msgstr "" |
| 201 |
"这是最坏的情况,因为它是单线程的,并且在之前的一个之后严格执行一个写操作,即I/O深度为1的运行。请查看<<s-latency-iops>>。" |
| 202 |
|
| 203 |
#. type: Plain text |
| 204 |
#: benchmark.adoc:100 |
| 205 |
msgid "" |
| 206 |
"Furthermore, see our <<ch-latency>> chapter for some typical performance " |
| 207 |
"values." |
| 208 |
msgstr "此外,有关一些典型的性能值,请参阅我们的<<ch-latency>>一章。" |