# HStream Admin
我们可以运行以下命令来使用 hstream admin：
docker run -it --rm --name some-hstream-admin --network host hstreamdb/hstream:latest hadmin --help
======= HStream Admin CLI =======
Usage: hadmin COMMAND
Available options:
-h,--help Show this help text
Available commands:
server Admin command
store Internal store admin command
1
2
3
4
5
6
7
8
9
10
11
2
3
4
5
6
7
8
9
10
11
为了便于演示说明，我们将在 hstream 容器中执行一个交互式 bash shell 来使用 hstream admin，
下面的示例用法是在你在快速启动中启动的集群上演示的，请相应调整。
docker exec -it docker_hserver_1 bash
$ hadmin server --help
1
2
2
Usage: hadmin server [--host SERVER-HOST] [--port INT]
[--log-level [critical|fatal|warning|info|debug]]
(COMMAND | COMMAND)
Admin command
Available options:
--host SERVER-HOST server host admin value (default: "127.0.0.1")
--port INT server admin port value (default: 6570)
--log-level [critical|fatal|warning|info|debug]
log level (default: info)
-h,--help Show this help text
Available commands:
sql Start an interactive SQL shell
stats Get the stats of an operation on a stream
stream Stream command
sub Subscription command
view View command
status Get the status of the HServer cluster
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
# 检查集群状态
> hadmin server status
+---------+---------+-------------------+
| node_id | state | address |
+---------+---------+-------------------+
| 100 | Running | 192.168.64.4:6570 |
| 101 | Running | 192.168.64.5:6572 |
+---------+---------+-------------------+
1
2
3
4
5
6
7
2
3
4
5
6
7
# 资源管理
# Streams
> hadmin server stream create --name s1
OK
> hadmin server stream list
+------+----------------------+
| name | replication_property |
+------+----------------------+
| s1 | node:3 |
+------+----------------------+
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
# Subscriptions
> hadmin server sub create --name sub1 --stream s1
OK
> hadmin server sub list
+------+-------------+---------+
| id | stream_name | timeout |
+------+-------------+---------+
| sub1 | s1 | 60 |
+------+-------------+---------+
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
# 数据统计 (HStream Stats)
hadmin server stats <stats_category> <stats_name>
1
- stream_counter
append_totalappend_failed
- stream
appendsorappend_in_bytesappend_in_recordappend_in_requestsappend_failed_requests
- subscription_counter
resend_records
- subscription
sendsorsend_out_bytessend_out_recordsacksoracknowledgementsrequest_messagesresponse_messages
更详细的解释请阅读 hstream metrics。
# 示例
hadmin server stats stream appends
1
+-------------+--------------+--------------+---------------+
| stream_name | appends_1min | appends_5min | appends_10min |
+-------------+--------------+--------------+---------------+
| s1 | 3570393 | 3570572 | 3570604 |
+-------------+--------------+--------------+---------------+
1
2
3
4
5
2
3
4
5
# 交互式 SQL shell
HAdmin 有一个交互式的 SQL shell，你可以在其中获得流的统计数据。
hadmin server sql
sql>
1
2
2
# SQL Shell
使用 show tables; 来获得从服务器上收集的所有 value tables 。它解释了所有存在的数据。
sql> show tables;
+-------------------------------+------------------------------------------+
| Table | Description |
+-------------------------------+------------------------------------------+
| streams | A table that lists the streams created |
| | in the cluster. |
+-------------------------------+------------------------------------------+
| | For each server node, reports the |
| server_append_request_latency | estimated percentiles latency of server |
| | append request |
+-------------------------------+------------------------------------------+
| | For each server node, reports the |
| server_append_latency | estimated percentiles latency of server |
| | appends |
+-------------------------------+------------------------------------------+
| append_total_counter | Total append requests server received. |
+-------------------------------+------------------------------------------+
| | For each server node, reports the |
| append_throughput | estimated per-stream append throughput |
| | over various time periods. |
+-------------------------------+------------------------------------------+
| | For each server node, reports the |
| subscription_throughput | estimated per-stream append throughput |
| | over various time periods. |
+-------------------------------+------------------------------------------+
| | For each server node, reports the |
| read_throughput | estimated per-stream read throughput |
| | over various time periods. |
+-------------------------------+------------------------------------------+
| append_failed_counter | Failed append requests. |
+-------------------------------+------------------------------------------+
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
describe streams;
+----------------------+--------+------------------------------------------+
| Column | Type | Description |
+----------------------+--------+------------------------------------------+
| node_id | int | Node ID this row is for. |
+----------------------+--------+------------------------------------------+
| name | string | The name of the stream. |
+----------------------+--------+------------------------------------------+
| replication_property | string | Replication property configured for this |
| | | stream. |
+----------------------+--------+------------------------------------------+
1
2
3
4
5
6
7
8
9
10
11
2
3
4
5
6
7
8
9
10
11
你可以直观地使用 select 来获取和计算你想要的数据。
sql> select * from streams;
+---------+------+----------------------+
| node_id | name | replication_property |
+---------+------+----------------------+
| 1 | demo | node:3 |
+---------+------+----------------------+
1
2
3
4
5
6
2
3
4
5
6
# 示例
找到过去10分钟内吞吐量最高的前 5 个 stream。
SELECT streams.name, sum(append_throughput.throughput_10min) AS total_throughput
FROM append_throughput
LEFT JOIN streams ON streams.name = append_throughput.stream_name
GROUP BY stream_name
ORDER BY total_throughput DESC
LIMIT 0, 5;
1
2
3
4
5
6
2
3
4
5
6