reduce sparse image sizes

Author: gingerwizard

docs/guides/best-practices/sparse-primary-indexes.md

@@ -331,7 +331,7 @@ ClickHouse is a <a href="https://clickhouse.com/docs/introduction/distinctive-fe
   - then by `URL`,
   - and lastly by `EventTime`:
 
-<Image img={sparsePrimaryIndexes01} size="lg" alt="Sparse Primary Indices 01" background="white"/>
+<Image img={sparsePrimaryIndexes01} size="md" alt="Sparse Primary Indices 01" background="white"/>
 
 `UserID.bin`, `URL.bin`, and `EventTime.bin` are the data files on disk where the values of the `UserID`, `URL`, and `EventTime` columns are stored.
 
@@ -355,7 +355,7 @@ Column values are not physically stored inside granules: granules are just a log
 The following diagram shows how the (column values of) 8.87 million rows of our table
 are organized into 1083 granules, as a result of the table's DDL statement containing the setting `index_granularity` (set to its default value of 8192).
 
-<Image img={sparsePrimaryIndexes02} size="lg" alt="Sparse Primary Indices 02" background="white"/>
+<Image img={sparsePrimaryIndexes02} size="md" alt="Sparse Primary Indices 02" background="white"/>
 
 The first (based on physical order on disk) 8192 rows (their column values) logically belong to granule 0, then the next 8192 rows (their column values) belong to granule 1 and so on.
 
@@ -390,7 +390,7 @@ For example
 
 In total the index has 1083 entries for our table with 8.87 million rows and 1083 granules:
 
-<Image img={sparsePrimaryIndexes03b} size="lg" alt="Sparse Primary Indices 03b" background="white"/>
+<Image img={sparsePrimaryIndexes03b} size="md" alt="Sparse Primary Indices 03b" background="white"/>
 
 :::note
 - For tables with [adaptive index granularity](/whats-new/changelog/2019.md/#experimental-features-1), there is also one "final" additional mark stored in the primary index that records the values of the primary key columns of the last table row, but because we disabled adaptive index granularity (in order to simplify the discussions in this guide, as well as make the diagrams and results reproducible), the index of our example table doesn't include this final mark.
@@ -615,7 +615,7 @@ We discuss that second stage in more detail in the following section.
 
 The following diagram illustrates a part of the primary index file for our table.
 
-<Image img={sparsePrimaryIndexes04} size="lg" alt="Sparse Primary Indices 04" background="white"/>
+<Image img={sparsePrimaryIndexes04} size="md" alt="Sparse Primary Indices 04" background="white"/>
 
 As discussed above, via a binary search over the index’s 1083 UserID marks, mark 176 was identified. Its corresponding granule 176 can therefore possibly contain rows with a UserID column value of 749.927.693.
 
@@ -637,7 +637,7 @@ In ClickHouse the physical locations of all granules for our table are stored in
 
 The following diagram shows the three mark files `UserID.mrk`, `URL.mrk`, and `EventTime.mrk` that store the physical locations of the granules for the table’s `UserID`, `URL`, and `EventTime` columns.
 
-<Image img={sparsePrimaryIndexes05} size="lg" alt="Sparse Primary Indices 05" background="white"/>
+<Image img={sparsePrimaryIndexes05} size="md" alt="Sparse Primary Indices 05" background="white"/>
 
 We have discussed how the primary index is a flat uncompressed array file (primary.idx), containing index marks that are numbered starting at 0.
 
@@ -688,7 +688,7 @@ The indirection provided by mark files avoids storing, directly within the prima
 
 The following diagram and the text below illustrate how for our example query ClickHouse locates granule 176 in the UserID.bin data file.
 
-<Image img={sparsePrimaryIndexes06} size="lg" alt="Sparse Primary Indices 06" background="white"/>
+<Image img={sparsePrimaryIndexes06} size="md" alt="Sparse Primary Indices 06" background="white"/>
 
 We discussed earlier in this guide that ClickHouse selected the primary index mark 176 and therefore granule 176 as possibly containing matching rows for our query.
 
@@ -802,7 +802,7 @@ We have marked the key column values for the first table rows for each granule i
 
 Suppose UserID had low cardinality. In this case it would be likely that the same UserID value is spread over multiple table rows and granules and therefore index marks. For index marks with the same UserID, the URL values for the index marks are sorted in ascending order (because the table rows are ordered first by UserID and then by URL). This allows efficient filtering as described below:
 
-<Image img={sparsePrimaryIndexes07} size="lg" alt="Sparse Primary Indices 06" background="white"/>
+<Image img={sparsePrimaryIndexes07} size="md" alt="Sparse Primary Indices 06" background="white"/>
 
 There are three different scenarios for the granule selection process for our abstract sample data in the diagram above:
 
@@ -816,7 +816,7 @@ There are three different scenarios for the granule selection process for our ab
 
 When the UserID has high cardinality then it is unlikely that the same UserID value is spread over multiple table rows and granules. This means the URL values for the index marks are not monotonically increasing:
 
-<Image img={sparsePrimaryIndexes08} size="lg" alt="Sparse Primary Indices 06" background="white"/>
+<Image img={sparsePrimaryIndexes08} size="md" alt="Sparse Primary Indices 06" background="white"/>
 
 As we can see in the diagram above, all shown marks whose URL values are smaller than W3 are getting selected for streaming its associated granule's rows into the ClickHouse engine.
 
@@ -850,7 +850,7 @@ ALTER TABLE hits_UserID_URL MATERIALIZE INDEX url_skipping_index;
 ```
 ClickHouse now created an additional index that is storing - per group of 4 consecutive [granules](#data-is-organized-into-granules-for-parallel-data-processing) (note the `GRANULARITY 4` clause in the `ALTER TABLE` statement above) - the minimum and maximum URL value:
 
-<Image img={sparsePrimaryIndexes13a} size="lg" alt="Sparse Primary Indices 13a" background="white"/>
+<Image img={sparsePrimaryIndexes13a} size="md" alt="Sparse Primary Indices 13a" background="white"/>
 
 The first index entry (‘mark 0’ in the diagram above) is storing the minimum and maximum URL values for the [rows belonging to the first 4 granules of our table](#data-is-organized-into-granules-for-parallel-data-processing).
 
@@ -945,11 +945,11 @@ OPTIMIZE TABLE hits_URL_UserID FINAL;
 
 Because we switched the order of the columns in the primary key, the inserted rows are now stored on disk in a different lexicographical order (compared to our [original table](#a-table-with-a-primary-key)) and therefore also the 1083 granules of that table are containing different values than before:
 
-<Image img={sparsePrimaryIndexes10} size="lg" alt="Sparse Primary Indices 10" background="white"/>
+<Image img={sparsePrimaryIndexes10} size="md" alt="Sparse Primary Indices 10" background="white"/>
 
 This is the resulting primary key:
 
-<Image img={sparsePrimaryIndexes11} size="lg" alt="Sparse Primary Indices 11" background="white"/>
+<Image img={sparsePrimaryIndexes11} size="md" alt="Sparse Primary Indices 11" background="white"/>
 
 That can now be used to significantly speed up the execution of our example query filtering on the URL column in order to calculate the top 10 users that most frequently clicked on the URL "http://public_search":
 ```sql
@@ -1097,7 +1097,7 @@ Ok.
 - if new rows are inserted into the source table hits_UserID_URL, then that rows are automatically also inserted into the implicitly created table
 - Effectively the implicitly created table has the same row order and primary index as the [secondary table that we created explicitly](/guides/best-practices/sparse-primary-indexes#option-1-secondary-tables):
 
-<Image img={sparsePrimaryIndexes12b1} size="lg" alt="Sparse Primary Indices 12b1" background="white"/>
+<Image img={sparsePrimaryIndexes12b1} size="md" alt="Sparse Primary Indices 12b1" background="white"/>
 
 ClickHouse is storing the [column data files](#data-is-stored-on-disk-ordered-by-primary-key-columns) (*.bin), the [mark files](#mark-files-are-used-for-locating-granules) (*.mrk2) and the [primary index](#the-primary-index-has-one-entry-per-granule) (primary.idx) of the implicitly created table in a special folder withing the ClickHouse server's data directory:
 
@@ -1181,7 +1181,7 @@ ALTER TABLE hits_UserID_URL
 - please note that projections do not make queries that use ORDER BY more efficient, even if the ORDER BY matches the projection's ORDER BY statement (see https://github.com/ClickHouse/ClickHouse/issues/47333)
 - Effectively the implicitly created hidden table has the same row order and primary index as the [secondary table that we created explicitly](/guides/best-practices/sparse-primary-indexes#option-1-secondary-tables):
 
-<Image img={sparsePrimaryIndexes12c1} size="lg" alt="Sparse Primary Indices 12c1" background="white"/>
+<Image img={sparsePrimaryIndexes12c1} size="md" alt="Sparse Primary Indices 12c1" background="white"/>
 
 ClickHouse is storing the [column data files](#data-is-stored-on-disk-ordered-by-primary-key-columns) (*.bin), the [mark files](#mark-files-are-used-for-locating-granules) (*.mrk2) and the [primary index](#the-primary-index-has-one-entry-per-granule) (primary.idx) of the hidden table in a special folder (marked in orange in the screenshot below) next to the source table's data files, mark files, and primary index files:
 
@@ -1449,7 +1449,7 @@ In the following we illustrate why it's beneficial for the compression ratio of
 
 The diagram below sketches the on-disk order of rows for a primary key where the key columns are ordered by cardinality in ascending order:
 
-<Image img={sparsePrimaryIndexes14a} size="lg" alt="Sparse Primary Indices 14a" background="white"/>
+<Image img={sparsePrimaryIndexes14a} size="md" alt="Sparse Primary Indices 14a" background="white"/>
 
 We discussed that [the table's row data is stored on disk ordered by primary key columns](#data-is-stored-on-disk-ordered-by-primary-key-columns).
 
@@ -1461,7 +1461,7 @@ and locality (the more similar the data is, the better the compression ratio is)
 
 In contrast to the diagram above, the diagram below sketches the on-disk order of rows for a primary key where the key columns are ordered by cardinality in descending order:
 
-<Image img={sparsePrimaryIndexes14b} size="lg" alt="Sparse Primary Indices 14b" background="white"/>
+<Image img={sparsePrimaryIndexes14b} size="md" alt="Sparse Primary Indices 14b" background="white"/>
 
 Now the table's rows are first ordered by their `ch` value, and rows that have the same `ch` value are ordered by their `cl` value.
 But because the first key column `ch` has high cardinality, it is unlikely that there are rows with the same `ch` value. And because of that is is also unlikely that `cl` values are ordered (locally - for rows with the same `ch` value).
@@ -1504,7 +1504,7 @@ The following diagram shows
 - the insert order of rows when the content changes (for example because of keystrokes typing the text into the text-area) and
 - the on-disk order of the data from the inserted rows when the `PRIMARY KEY (hash)` is used:
 
-<Image img={sparsePrimaryIndexes15a} size="lg" alt="Sparse Primary Indices 15a" background="white"/>
+<Image img={sparsePrimaryIndexes15a} size="md" alt="Sparse Primary Indices 15a" background="white"/>
 
 Because the `hash` column is used as the primary key column
 - specific rows can be retrieved [very quickly](#the-primary-index-is-used-for-selecting-granules), but
@@ -1519,7 +1519,7 @@ The following diagram shows
 - the insert order of rows when the content changes (for example because of keystrokes typing the text into the text-area) and
 - the on-disk order of the data from the inserted rows when the compound `PRIMARY KEY (fingerprint, hash)` is used:
 
-<Image img={sparsePrimaryIndexes15b} size="lg" alt="Sparse Primary Indices 15b" background="white"/>
+<Image img={sparsePrimaryIndexes15b} size="md" alt="Sparse Primary Indices 15b" background="white"/>
 
 Now the rows on disk are first ordered by `fingerprint`, and for rows with the same fingerprint value, their `hash` value determines the final order.