it takes too long to describe a tag after creating it under wsl docker #435
Comments
|
This problem is due the synchronization of the meta infos among our different processes. For now the meta infos are updated through periodic polling. To address this problem we need to some kind of broadcast mechanism or forced update through a user command. |
|
This is definitely a bug. When we DESCRIBE a tag/edge/etc., we should read it from the meta server |
|
It is a bug. Currently, show space/edge/tag and describe space are fetched from meta server, but describe tag/edge is fetched from meta client cache. |
|
@laura-ding Do you want to fix it? |
|
My question is: why 10 seconds? The fetch loop cycle is only 1 or 2 sec. |
|
I thought the user created a schema, usually read back the schema firstly, then insert the data via the cached schema. So describe uses the cached schema, it can tell the user in advance whether you can insert data. |
|
@dangleptr I will modify it to get schema by meta server |
|
Under docker, the default value of load_data_interval_secs is 2min, now all describe command unified get data from metad, it doesn't need to wait. |
#### What type of PR is this?
- [ ] bug
- [x] feature
- [ ] enhancement
#### What does this PR do?
#### Which issue(s)/PR(s) this PR relates to?
#### Special notes for your reviewer, ex. impact of this fix, etc:
#### Additional context/ Design document:
#### Checklist:
- [x] Documentation affected (Please add the label if documentation needs to be modified.)
- [x] Incompatibility (If it breaks the compatibility, please describe it and add the corresponding label.）
- [ ] If it's needed to cherry-pick (If cherry-pick to some branches is required, please label the destination version(s).)
- [ ] Performance impacted: Consumes more CPU/Memory
#### Release notes:
A little incompatible: The returned column names are the same sequence with the declared columns, while it is random in former release.
```
MATCH (v:player)-[:like]->(v2)
WHERE v.name == "Tony Parker" and v2.age == 42
WITH *, v.age + 100 AS age
RETURN *, v2.name
```
Now this query returns:
```
| v | v2 | age | v2.name |
```
New features:
```
MATCH (m)-[]-(n), (n)-[]-(l) WHERE id(m)=="Tim Duncan"
RETURN m.name AS n1, n.name AS n2, l.name AS n3 ORDER BY n1, n2, n3 LIMIT 10
```
```
MATCH (m)-[]-(n) WHERE id(m)=="Tim Duncan"
MATCH (n)-[]-(l)
RETURN m.name AS n1, n.name AS n2, l.name AS n3 ORDER BY n1, n2, n3 LIMIT 10
```
```
MATCH (m)-[]-(n) WHERE id(m)=="Tim Duncan"
OPTIONAL MATCH (n)<-[:serve]-(l)
RETURN m.name AS n1, n.name AS n2, l AS n3 ORDER BY n1, n2, n3 LIMIT 10
```
We have more test cases in pr vesoft-inc#3537
Migrated from vesoft-inc#3519
Co-authored-by: cpw <13495049+CPWstatic@users.noreply.github.com>
My environment is wsl.
first I create a tag: CREATE TAG player4(name string, age int);
then i have to wait quite a long time ( about 10sec) to successfully describe tag player4.
During this long period, what i can get is [ERROR (-8)]: Schema not found for tag `player4'
By the way, I know the flag load_data_interval_secs , but i can't find it in the config files in the docker any more.
The text was updated successfully, but these errors were encountered: