Bug 1341709

Summary: Unable to install feeipa server on ppc64le f24 VM
Product: [Fedora] Fedora Reporter: Menanteau Guy <menantea>
Component: freeipaAssignee: IPA Maintainers <ipa-maint>
Status: CLOSED ERRATA QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 24CC: abokovoy, awilliam, ipa-maint, jhrozek, menantea, mkosek, nhosoi, pvoborni, rcritten, ssorce, wibrown
Target Milestone: ---   
Target Release: ---   
Hardware: ppc64le   
OS: Linux   
Whiteboard:
Fixed In Version: 389-ds-base-1.3.5.5-1.el7 Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-03-15 16:50:19 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1320886    
Attachments:
Description Flags
ipa server log
none
cat /proc/self/status
none
cat /proc/meminfo none

Description Menanteau Guy 2016-06-01 14:37:16 UTC
I tried to set up a freeipa server on a ppc64le f24 server VM using the command 
"ipa-server-install" without any success.
I used a VM installed with following iso:
https://dl.fedoraproject.org/pub/fedora-secondary/releases/test/24_Beta/Server/ppc64/iso/Fedora-Server-dvd-ppc64-24_Beta-1.6.iso

Hereafter is end of console trace:
The following operations may take some minutes to complete.
Please wait until the prompt is returned.

Configuring NTP daemon (ntpd)
  [1/4]: stopping ntpd
  [2/4]: writing configuration
  [3/4]: configuring ntpd to start on boot
  [4/4]: starting ntpd
Done configuring NTP daemon (ntpd).
Configuring directory server (dirsrv). Estimated time: 1 minute
  [1/46]: creating directory server user
  [2/46]: creating directory server instance
  [error] RuntimeError: failed to create ds instance Command '/usr/sbin/setup-ds.pl --silent --logfile - -f /tmp/tmpFqTqun' returned non-zero exit status 1
ipa.ipapython.install.cli.install_tool(Server): ERROR    failed to create ds instance Command '/usr/sbin/setup-ds.pl --silent --logfile - -f /tmp/tmpFqTqun' returned non-zero exit status 1
ipa.ipapython.install.cli.install_tool(Server): ERROR    The ipa-server-install command failed. See /var/log/ipaserver-install.log for more information


I see traces related to memory in the ipaserver-install.log but I am not sure it is the cause of my problem. Anyway I tried to run my VM machine with more 
RAM (8GB) and a bigger swap (4G) but it did not change the result.

Comment 1 Menanteau Guy 2016-06-01 14:38:02 UTC
Created attachment 1163702 [details]
ipa server log

Comment 2 Alexander Bokovoy 2016-06-01 14:54:21 UTC
This looks like a lack of memory:

2016-05-31T12:28:19Z DEBUG stdout=[16/05/31:08:28:19] - [Setup] Info Could not import LDIF file '/var/lib/dirsrv/boot.ldif'.  Error: 3072.  Output: importing data ...
[31/May/2016:08:28:19.525709246 -0400] WARNING: Import is running with nsslapd-db-private-import-mem on; No other process is allowed to access the database
[31/May/2016:08:28:19.550369356 -0400] check_and_set_import_cache: pagesize: 65536, pages: 1144, procpages: 497
[31/May/2016:08:28:19.551091971 -0400] After allocating import cache 29248KB, the available memory is 43968KB, which is less than the hard limit 102400KB. Please decrease the import cache size and rerun import.
[31/May/2016:08:28:19.551793917 -0400] dblayer_instance_start fail: Cannot allocate memory (12)

Comment 3 Alexander Bokovoy 2016-06-01 14:56:23 UTC
William, any idea why 389-ds freaks out on import?

Comment 4 wibrown@redhat.com 2016-06-02 00:49:51 UTC
https://fedorahosted.org/389/ticket/48846

This has some fixes around memory size checks related to the import, and may affect ppc64.

1.3.5.5 will likely contain the fix you need. 48846 will be expanded to take this into account. I think the the issue is the same as on a 32 bit platform, where vmsize is set low on ppc64, unlike x86_64 where it's set high, and basically discarded in our util_info_sys_pages() check. 

It's very likely that I'll be removing from the vmsize from consideration of util_info_sys_pages() for some other reasons too. 

I'd really like to see the output of 'cat /proc/self/status' and 'cat /proc/meminfo' of the affected system, just to be sure of a few details if possible.

Comment 5 wibrown@redhat.com 2016-06-02 01:08:31 UTC
https://fedorahosted.org/389/ticket/48863

Comment 6 Menanteau Guy 2016-06-02 07:20:39 UTC
Created attachment 1163931 [details]
cat /proc/self/status

Comment 7 Menanteau Guy 2016-06-02 07:22:13 UTC
Created attachment 1163933 [details]
cat /proc/meminfo

here after is the qemu command I used with 8G:
/usr/bin/qemu-system-ppc64 -name /home/menantea/VM/f24_91 -enable-kvm -M pseries -smp 4 -m 8G -nographic -nodefaults -monitor stdio -se
rial pty -device virtio-net-pci,netdev=net10131,mac=c0:ff:ff:00:00:91 -netdev br
idge,br=br0,id=net10131 -cdrom isobe_beta1.6 -drive file=hd1.qcow2 -drive file=h
d2.qcow2 -S

Comment 8 wibrown@redhat.com 2016-06-02 22:59:09 UTC
Thank you for that information. I'll keep it in mind while testing the fix. I will provide status updates as they become available.

Comment 9 Menanteau Guy 2016-06-17 16:04:31 UTC
I didn't end all tests I wanted to do but at least the new 389-ds-base-1.3.5.5-1
resolved my problem on f24 ppc64le and the "ipa-server-install" command go to the end without error. Thanks for the fix.

Comment 10 Adam Williamson 2017-03-15 16:50:19 UTC
Per Guy's comment, this seems to be fixed. F24 is up to 1.3.5.15-1 at this point.