-
Type: Bug
-
Resolution: Unresolved
-
Priority: Minor - P4
-
None
-
Affects Version/s: 3.4.21, 4.0.9, 3.6.13
-
Component/s: mongodump, mongorestore
-
None
-
1,918
UPDATE:
Tools on windows connecting to localhost suffer a long name resolution delay. A workaround is to connect to 127.0.0.1.
ORIGINAL:
After migration from MongoDB v2.0.9 to v4.0.9 the main bad thing I noticed is a huge time for execution dump/restore commands.
For v2.0.9 those operations were almost instantaneous (dataset is tiny). But same commands for same datasets with new binaries for v4.0.9 take seconds for the same operations.
From the verbose log, I can see that 2 seconds it does nothing before starting to perform the operation.
2019-06-04T11:14:26.413+0300 will listen for SIGTERM, SIGINT, and SIGKILL 2019-06-04T11:14:28.430+0300 enqueued collection 'potrust_1.ReadStats' ... 2019-06-04T11:14:29.503+0300 finishing dump
{{}}
I use dump/restore a lot in test cases between tests (actually hundreds of such operations) so now my tests run like 2-3 times longer (
Am I doing something wrong or is it really should be so slow now? Is it because tools are now written in go?
- depends on
-
GODRIVER-1196 Slow localhost connection on Windows
- Closed
- related to
-
PYTHON-356 Windows localhost connection slowdowns.
- Closed
- links to