View Issue Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0001294 | bareos-core | file daemon | public | 2020-12-19 20:39 | 2023-09-05 14:01 |
Reporter | bsperduto | Assigned To | bruno-at-bareos | ||
Priority | normal | Severity | minor | Reproducibility | always |
Status | feedback | Resolution | open | ||
Platform | Linux | OS | CentOS | OS Version | 8 |
Product Version | 20.0.0 | ||||
Summary | 0001294: LibCloud Accurate Files Memory Leak | ||||
Description | Hi, In the current LibCloud with an accurate backup configuration has a frequent tendency to create an out of space error. In an accurate backup mode files that are determined to be unchanged are still passed up to the worker queue to be downloaded (bareos/core/src/plugins/filed/python/libcloud/bareos_libcloud_api/bucket_explorer.py L141). When the file makes its way through the queue and through start_backup_file it appears the director never triggers any file IOs against this file and never calls IO_CLOSE on the file. This will generate a memory leak on large buckets as these files are never purged until the job finishes and jobs are frequently aborted due to out of memory. I have a few thoughts on how to work around this but I'd like some input before proceeding. | ||||
Tags | No tags attached. | ||||
Hello, cleaning up the bugs entries. May we ask you if this is still reproducible with recent version like 22.1.0 ? If yes are you able to share some configuration to implement a systemtest that can reproduce the problem. Of course, if you know how to, please open a PR on the github. |
|
Date Modified | Username | Field | Change |
---|---|---|---|
2020-12-19 20:39 | bsperduto | New Issue | |
2023-09-05 14:01 | bruno-at-bareos | Assigned To | => bruno-at-bareos |
2023-09-05 14:01 | bruno-at-bareos | Status | new => feedback |
2023-09-05 14:01 | bruno-at-bareos | Note Added: 0005376 |