Mastering Redis: Take your knowledge of Redis to the next level to build enthralling applications with ease

Good

Good content but horrible editing. Let me quickly grom the state of redis clustering options.

Feb/2019. Picked this up (ebook from packt site) as it was the newest publication listed at the Redis site. So far, I'm on chapter 3 and have found numerous errata and confusing examples. The book clearly contains some useful information, but I keep finding myself thinking "wait, that can't be right". I've found errors in the coding examples and errors in the text that seem to go beyond simple typo and make me question the author's expertise on the subject matter.Example1: In chapter 3, on memory management, the author is discussing the "rdbchecksumdirective" and indicates that this enables a CRC64 checksum to be added to the end of a RDB snapshot. The author writes: "Performing an RDB snapshot with this CRC64 checksum imposes a 10% increase in memory usage when Redis spawns a child process that saves the snapshot in the disk." ... wait, that can't be right ... why would memory usage increase by 10% when computing a checksum??? OK, so I looked this up in the redis config file. First, I did a google search for "rdbchecksumdirective" and only found hits related to the book. Why? Because it's actually "rdbchecksum" (ok, simple typo there). The actual redis.conf file indicates "there is a performance hit to pay (around 10%) when saving and loading RDB files, so you can disable it for maximum performances" - this makes total sense. There is a CPU performance penalty for computing a checksum - not an increase in memory usage as indicated by the book.Example2: In chapter 2, there is a discussion of the HyperLogLog type. [This is like a "set", but it doesn't actually store the members of the set, but is able to estimate the number of members (cardinality) without actually storing the members.] The author compares the cardinality results between using an actual Set and using a HyperLogLog. For a regular set, he adds elements 1..52111. Then gets cardinality showing 52411 (typo?) but then he compares this incorrect number to the estimated cardinality from the HyperLogLog of 52213. At this point, I don't know if the resulting math makes any sense at all being based on incorrect numbers.Example3: In a (simple?) diagram of a bitmap, he shows 8 bits in a byte, but then shows the Redis commands only using 1 (leftmost) bit per byte. Setting bit "1" (0-based) is shown to set the 1st bit in the 2nd byte. Bit "2" is shown to be the 1st bit in the 3rd byte. ... wait, this can't be right??? In the text, this is continued with and example setting bit 2445 where the author refers to it as the "2445 byte".So, I'd say this is a good book to get you going, but be prepared to fact check what you are reading.EDIT: Now reading Chapter 4 and need to lower to 2 stars. Continuing to find errors/issues.Chapter 4 starts diving into some of the C source code for Redis. In one paragraph, the author sees fit to draw our attention to 3 #defines which are internal states of the "AOF" mechanism - yet nowhere in the entire book does the author talk about AOF. I have no idea why the author would tell us that "REDIS_AOF_WAIT_REWRITE" is defined as 2. Basically: "See... there are these #defines... and this one is defined as 2!" ?And in the next paragraph... "redis.h defines nineteen different flags to set a bit to 1 using the << bitwise left-shift operator so that a client information can be represented in a single byte." 1) Why are we talking about some internally defined constants? 2) Why is the author explaining a bitwise left-shift operator? 3) How has the author managed to fit 19 flags into a byte?