Reduce image size to speed up loading times

Author: wangbin579

Chapter1.md

@@ -52,7 +52,7 @@ The Paxos protocol relies on a majority-based mechanism rooted in set theory to
 
 Raft, another widely used protocol, is essentially a simplified version of the Paxos protocol. The following section describes Meta's implementation of MySQL high availability using the Raft protocol [38], which simplifies traditional high availability processes.
 
-<img src="media/91bc13de07d658b009ae498b5afa4ac1.png" style="zoom: 50%;" />
+<img src="media/91bc13de07d658b009ae498b5afa4ac1.gif" style="zoom: 50%;" />
 
 Figure 1-5. Reasons Meta uses raft.
 
@@ -100,13 +100,13 @@ During TPC-C data loading tests with BenchmarkSQL under normal pressure, the MyS
 
 Here is a partial screenshot of the error log from the MySQL primary:
 
-![](media/4f558a475444869186a0641f934db534.png)
+![](media/4f558a475444869186a0641f934db534.gif)
 
 Figure 1-10. A partial screenshot of the error log from the MySQL primary.
 
 Here is a partial screenshot of the error log from the MySQL secondary:
 
-![](media/974f0f1b091afaba18ea375a73d87441.png)
+![](media/974f0f1b091afaba18ea375a73d87441.gif)
 
 Figure 1-11. A partial screenshot of the error log from the MySQL secondary.
 

Chapter10.md

@@ -155,7 +155,7 @@ In MySQL asynchronous replication, the secondary has two threads: the IO thread
 
 For example, during Group Replication secondary replay, one of the main bottlenecks captured is the SQL thread, specifically seen in the *rpl_rca_sql-0* thread as shown in the figure below.
 
-![](media/3b54c6545f0c49be73fd3afd9c6ffe79.png)
+![](media/3b54c6545f0c49be73fd3afd9c6ffe79.gif)
 
 Figure 10-7. Top two CPU-consuming threads in Group Replication secondary replay.
 
@@ -312,7 +312,7 @@ If two transactions occurring close together in time modify the same data in MyS
 
 To better understand this problem, consider the following case:
 
-![](media/9aa41ee1e12bb8a72dd240dc42b44ba5.png)
+![](media/9aa41ee1e12bb8a72dd240dc42b44ba5.gif)
 
 Figure 10-15. Typical examples of *sequence_number* and *last_committed*.
 
@@ -352,13 +352,13 @@ The decision to disable NUMA should be made based on a thorough analysis. Testin
 
 Based on MySQL's Performance Schema, statistics on latch operations are gathered. See the results in the following figure:
 
-![](media/08fb8a25d9542121b1e5b5ccb3cbdf6f.png)
+![](media/08fb8a25d9542121b1e5b5ccb3cbdf6f.gif)
 
 Figure 10-19. The bottleneck of pending_jobs_lock has been exposed in the Performance Schema.
 
 From the figure, it can be seen that in a NUMA environment, *Relay_log_info::pending_jobs_lock* ranks fifth in importance in relation to replay. Binding the MySQL secondary instance to NUMA node 0, the test evaluates whether the Performance Schema statistics change under an SMP environment. The specific test results are shown in the following figure:
 
-![](media/28708a58742f84033b84ff551cda3b1b.png)
+![](media/28708a58742f84033b84ff551cda3b1b.gif)
 
 Figure 10-20. The bottleneck of pending_jobs_lock has been alleviated in an SMP environment.
 

Chapter3.md

@@ -119,7 +119,7 @@ Figure 3-3. Description of the official worklog for redo log optimization.
 
 To address the scalability problems of redo log, developers have subdivided this problem into the following sub-problems:
 
-![](media/5502725fcc5648f2f6d18aaf68832505.png)
+![](media/5502725fcc5648f2f6d18aaf68832505.gif)
 
 Figure 3-4. Requirements of the official worklog for redo log optimization.
 

Chapter4_7.md

@@ -45,7 +45,7 @@ To address these problems, additional tools like packet capture can help determi
 
 The following figure depicts the classic TCP state machine used for transitions between TCP states.
 
-![](media/ff831787f6acc1e38a7f70dc47925ee0.png)
+![](media/ff831787f6acc1e38a7f70dc47925ee0.gif)
 
 Figure 4-35. Classic TCP state machine overview.
 
@@ -112,7 +112,7 @@ From the figure, it is evident that the delay between two Paxos instances is aro
 
 In general, the test conclusions of pipelining and batching are consistent with the conclusions in the following paper [48]:
 
-![](media/73dfb3c0a76ecd9d6faa62c025ace5e4.png)
+![](media/73dfb3c0a76ecd9d6faa62c025ace5e4.gif)
 
 Figure 4-42. Impact of pipelining and batching on Paxos performance.
 
@@ -168,7 +168,7 @@ For more advanced features, refer to the *tcpdump* documentation for detailed co
 
 Many software professionals lack in-depth knowledge of TCP/IP logic reasoning, which often leads to misidentifying problems as mysterious problems. Some are discouraged by the complexity of TCP/IP networking literature, while others are misled by confusing details in Wireshark. For instance, a DBA facing performance problems might misinterpret packet capture data in Wireshark, erroneously concluding that TCP retransmissions are the cause.
 
-![](media/3a5d10caf25003e7ea4dcace59a181f6.png)
+![](media/3a5d10caf25003e7ea4dcace59a181f6.gif)
 
 Figure 4-45. Packet capture screenshot provided by DBA suspecting retransmission problems.
 

Chapter7.md

@@ -309,7 +309,7 @@ Comparative analysis shows that the CATS algorithm significantly reduces deadloc
 
 The above test results align closely with MySQL's official findings. The following two figures, based on official tests [57], demonstrate the significant effectiveness of the CATS algorithm.
 
-![](media/853d21533f748c1c56a4151869a82a27.png)
+![](media/853d21533f748c1c56a4151869a82a27.gif)
 
 Figure 7-13. Comparison of CATS and FIFO in TPS and mean latency: insights from the MySQL blog.
 

Chapter8.md

@@ -65,7 +65,7 @@ From the figure, it can be seen that the number of rows is still large, indicati
 
 Two different debug versions of MySQL were compiled: one with anomalies and one normal. Debug versions were used to capture useful function call relationships through debug traces. When executing the problematic SQL statement on the version with anomalies, the relevant debug trace information is as follows:
 
-![](media/149ca62fd014bd12b60c77573a49757d.png)
+![](media/149ca62fd014bd12b60c77573a49757d.gif)
 
 Figure 8-6. Debug trace information for the abnormal version.
 

Chapter9.md

@@ -198,7 +198,7 @@ When there are no theoretical problems with the Mencius algorithm, introducing a
 
 First, let's investigate the performance problems of the MySQL Mencius algorithm implementation. The following figure illustrates the network interaction status when the Mencius algorithm operates stably with a network delay of 10ms:
 
-![](media/77d0c0bdc5ce8574c6ad319864abb032.png)
+![](media/77d0c0bdc5ce8574c6ad319864abb032.gif)
 
 Figure 9-6. Insights into the Mencius protocol from packet capture data.