Note: This is a beta release of Red Hat Bugzilla 5.0. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Also email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback here.
Bug 1597384 - Async operations can hang when the server is running nunc-stans
Summary: Async operations can hang when the server is running nunc-stans
Alias: None
Product: Red Hat Enterprise Linux 7
Classification: Red Hat
Component: 389-ds-base
Version: 7.4
Hardware: All
OS: Linux
Target Milestone: rc
: ---
Assignee: thierry bordaz
QA Contact: RHDS QE
: 1600764 (view as bug list)
Depends On:
Blocks: 1597530
TreeView+ depends on / blocked
Reported: 2018-07-02 20:02 UTC by mreynolds
Modified: 2018-10-30 12:41 UTC (History)
10 users (show)

Fixed In Version: 389-ds-base-
Doc Type: No Doc Update
Doc Text:
Clone Of:
: 1597530 (view as bug list)
Last Closed: 2018-10-30 10:14:34 UTC
Target Upstream Version:

Attachments (Terms of Use)

System ID Priority Status Summary Last Updated
Red Hat Knowledge Base (Solution) 3553761 None None None 2018-08-07 09:57:22 UTC
Red Hat Product Errata RHSA-2018:3127 None None None 2018-10-30 10:15:20 UTC

Description mreynolds 2018-07-02 20:02:14 UTC
This bug is created as a clone of upstream ticket:

#### Issue Description
This bug is a side effect of

With async operations on the same connection, the connection can be set several times in read_ready. This is done with connection_threadmain scheduling ns_handle_pr_read_ready (connection_make_readable_nolock).

When there is no more activity on the connection (no new req, timeout, closure..) and the server is listening to it (ns_handle_pr_read_ready), then if an async operation tries to schedule ns_handle_pr_read_ready it will hang (before completing the operation).

The consequence is one or more hanging operations on the connection with a stack like

    Thread 63 (Thread 0x7f794a711700 (LWP 28798)):
    #0  0x00007f7995e75c93 in select () from /lib64/
    #1  0x00007f7998e8046b in DS_Sleep (ticks=100) at 389-ds-base/ldap/servers/slapd/util.c:1086
    #2  0x0000000000438a6a in ns_connection_post_io_or_closing (conn=0x7f79550fc800) at 389-ds-base/ldap/servers/slapd/daemon.c:1835
    #3  0x0000000000427adc in connection_make_readable_nolock (conn=0x7f79550fc800) at 389-ds-base/ldap/servers/slapd/connection.c:1361                                                                                              
    #4  0x0000000000429e91 in connection_threadmain () at 389-ds-base/ldap/servers/slapd/connection.c:1724
    #5  0x00007f7996a0907b in _pt_root () from /lib64/
    #6  0x00007f79965ab36d in start_thread () from /lib64/
    #7  0x00007f7995e7fb4f in clone () from /lib64/      

A way to mitigate the problem is to set a idletimeout

#### Package Version and Platform
1.3.7, 1.3.8 and master

#### Steps to reproduce

test case is not systematic.
On rapid machine, it hangs one out of ten

#### Actual results
Some operations are not completing (hanging)

#### Expected results
Should not hang

Comment 6 German Parente 2018-08-01 16:08:11 UTC
*** Bug 1600764 has been marked as a duplicate of this bug. ***

Comment 7 Viktor Ashirov 2018-08-30 13:22:56 UTC
Build tested: 389-ds-base-

Cmocka unit tests pass:

PASS: test_nuncstans
PASS: test_nuncstans_stress_small
PASS: test_nuncstans_stress_large

No hang was observed during the acceptance testing with nunc-stans. NS is also disabled by default on new installs (bz1614501), so marking as VERIFIED, SanityOnly.

Comment 14 errata-xmlrpc 2018-10-30 10:14:34 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

Note You need to log in before you can comment on or make changes to this bug.