|
1 | | -0) Design |
| 1 | +# pg_dtm |
2 | 2 |
|
3 | | -General concept oberview. wiki link. This repo implements Snapshot Sharing mechnism. |
4 | | -Protocol description. README.protocol |
5 | | -Presentation |
| 3 | +### Design |
6 | 4 |
|
7 | | -1) Installing |
| 5 | +This repo implements distributed transaction manager using Snapshot Sharing mechanism. General concepts and alternative approaches described in postgres wiki https://wiki.postgresql.org/wiki/DTM. |
8 | 6 |
|
9 | | -* patch postgres |
10 | | -* install extension |
11 | | -* configure two postgreses |
12 | | -* run dtmd |
13 | | -* run postgreses |
| 7 | +Backend-DTM protocol description can be found in [dtmd/README](dtmd/README). |
14 | 8 |
|
15 | | -1b) Automatic provisioning |
| 9 | +### Installation |
16 | 10 |
|
17 | | -* For a wide deploy we use ansible. Layouts/Farms. More details later. |
18 | | - |
19 | | -2) Usage |
| 11 | +* Patch postgres using xtm.patch. After that build and install postgres in usual way. |
| 12 | +```bash |
| 13 | +cd ~/code/postgres |
| 14 | +patch -p1 < ~/code/pg_dtm/xtm.patch |
| 15 | +``` |
| 16 | +* Install pg_dtm extension. |
| 17 | +```bash |
| 18 | +export PATH=/path/to/pgsql/bin/:$PATH |
| 19 | +cd ~/code/pg_dtm |
| 20 | +make && make install |
| 21 | +``` |
| 22 | +* Run dtmd. |
| 23 | +```bash |
| 24 | +cd ~/code/pg_dtm/dtmd |
| 25 | +make |
| 26 | +./bin/dtmd & |
| 27 | +``` |
| 28 | +* To run something meaningful you need at leat two postgres instances. Also pg_dtm requires |
| 29 | +```bash |
| 30 | +initdb -D ./install/data1 |
| 31 | +initdb -D ./install/data2 |
| 32 | +echo "port = 5433" >> ./install/data2/postgresql.conf |
| 33 | +echo "shared_preload_libraries = 'pg_dtm'" >> ./install/data1/postgresql.conf |
| 34 | +echo "shared_preload_libraries = 'pg_dtm'" >> ./install/data2/postgresql.conf |
| 35 | +pg_ctl -D ./install/data1 -l ./install/data1/log start |
| 36 | +pg_ctl -D ./install/data2 -l ./install/data2/log start |
| 37 | +``` |
20 | 38 |
|
21 | | -now you can use global tx between this two nodes |
| 39 | +#### Automatic provisioning |
22 | 40 |
|
23 | | -table with two columns |
24 | | -```sql |
25 | | -example |
26 | | -``` |
| 41 | +For a cluster-wide deploy we use ansible, more details in tests/deploy_layouts. (Ansible instructions will be later) |
27 | 42 |
|
28 | | -3) Consistency testing |
| 43 | +### Usage |
29 | 44 |
|
30 | | -To ensure consistency we use simple bank test: perform a lot of simultaneous transfers between accounts on different servers, while constantly checking total amount of money on all accounts. |
| 45 | +Now cluster is running and you can use global tx between two nodes. |
31 | 46 |
|
32 | | -go run ... |
33 | 47 |
|
34 | | -also there is the test for measuring select performance. |
| 48 | +```sql |
| 49 | +create extension pg_dtm; -- node1 |
| 50 | + create extension pg_dtm; -- node2 |
| 51 | +select dtm_begin_transaction(); -- node1, returns global xid, e.g. 42 |
| 52 | + select dtm_join_transaction(42); -- node2, join global tx |
| 53 | +begin; -- node1 |
| 54 | + begin; -- node2 |
| 55 | +update accounts set amount=amount-100 where user_id=1; -- node1, transfer money from user#1 |
| 56 | + update accounts set amount=amount+100 where user_id=2; -- node2, to user#2 |
| 57 | +commit; -- node1 |
| 58 | + commit; -- node2 |
| 59 | +``` |
35 | 60 |
|
36 | | -4) Using with fdw. |
| 61 | +### Consistency testing |
| 62 | + |
| 63 | +To ensure consistency we use simple bank test: perform a lot of simultaneous transfers between accounts on different servers, while constantly checking total amount of money on all accounts. This test can be found in tests/perf. |
| 64 | + |
| 65 | +```bash |
| 66 | +> go run ./perf/* |
| 67 | + -C value |
| 68 | + Connection string (repeat for multiple connections) |
| 69 | + -a int |
| 70 | + The number of bank accounts (default 100000) |
| 71 | + -b string |
| 72 | + Backend to use. Possible optinos: transfers, fdw, pgshard, readers. (default "transfers") |
| 73 | + -g Use DTM to keep global consistency |
| 74 | + -i Init database |
| 75 | + -l Use 'repeatable read' isolation level instead of 'read committed' |
| 76 | + -n int |
| 77 | + The number updates each writer (reader in case of Reades backend) performs (default 10000) |
| 78 | + -p Use parallel execs |
| 79 | + -r int |
| 80 | + The number of readers (default 1) |
| 81 | + -s int |
| 82 | + StartID. Script will update rows starting from this value |
| 83 | + -v Show progress and other stuff for mortals |
| 84 | + -w int |
| 85 | + The number of writers (default 8) |
| 86 | +``` |
37 | 87 |
|
38 | | -patch |
39 | | -go run ... |
| 88 | +So previous installation can be initialized with: |
| 89 | +``` |
| 90 | +go run ./perf/*.go \ |
| 91 | +-C "dbname=postgres port=5432" \ |
| 92 | +-C "dbname=postgres port=5433" \ |
| 93 | +-g -i |
| 94 | +``` |
| 95 | +and tested with: |
| 96 | +``` |
| 97 | +go run ./perf/*.go \ |
| 98 | +-C "dbname=postgres port=5432" \ |
| 99 | +-C "dbname=postgres port=5433" \ |
| 100 | +-g |
| 101 | +``` |
40 | 102 |
|
41 | | -5) Using with pg_shard |
| 103 | +### Using with postres_fdw. |
42 | 104 |
|
43 | | -checkout repo |
44 | | -go run ... |
| 105 | +We also provide a patch, that enables support of global transactions with postres_fdw. After patching and installing postres_fdw it is possible to run same test via fdw usig key ```-b fdw```. |
45 | 106 |
|
46 | | -6) Results |
| 107 | +### Using with pg_shard |
47 | 108 |
|
48 | | -Some graphs |
| 109 | +Citus Data have branch in their pg_shard repo, that interacts with transaction manager. https://github.com/citusdata/pg_shard/tree/transaction_manager_integration |
| 110 | +To use this feature one should have following line in postgresql.conf (or set it via GUC) |
| 111 | +``` |
| 112 | +pg_shard.use_dtm_transactions = 1 |
| 113 | +``` |
0 commit comments