RHEL Engineering is moving the tracking of its product development work on RHEL 6 through RHEL 9 to Red Hat Jira (issues.redhat.com). If you're a Red Hat customer, please continue to file support cases via the Red Hat customer portal. If you're not, please head to the "RHEL project" in Red Hat Jira and file new tickets here. Individual Bugzilla bugs in the statuses "NEW", "ASSIGNED", and "POST" are being migrated throughout September 2023. Bugs of Red Hat partners with an assigned Engineering Partner Manager (EPM) are migrated in late September as per pre-agreed dates. Bugs against components "kernel", "kernel-rt", and "kpatch" are only migrated if still in "NEW" or "ASSIGNED". If you cannot log in to RH Jira, please consult article #7032570. That failing, please send an e-mail to the RH Jira admins at rh-issues@redhat.com to troubleshoot your issue as a user management inquiry. The email creates a ServiceNow ticket with Red Hat. Individual Bugzilla bugs that are migrated will be moved to status "CLOSED", resolution "MIGRATED", and set with "MigratedToJIRA" in "Keywords". The link to the successor Jira issue will be found under "Links", have a little "two-footprint" icon next to it, and direct you to the "RHEL project" in Red Hat Jira (issue links are of type "https://issues.redhat.com/browse/RHEL-XXXX", where "X" is a digit). This same link will be available in a blue banner at the top of the page informing you that that bug has been migrated.
Bug 1262143 - VM startup is very slow with large amounts of hotpluggable memory
Summary: VM startup is very slow with large amounts of hotpluggable memory
Alias: None
Product: Red Hat Enterprise Linux 7
Classification: Red Hat
Component: qemu-kvm-rhev
Version: 7.2
Hardware: ppc64le
OS: Linux
Target Milestone: rc
: 7.3
Assignee: David Gibson
QA Contact: Virtualization Bugs
Depends On: 1263039
Blocks: RHEV3.6PPC 1261812 1277183 1277184 1277186 RHEV3.6PPC_PCI_Passthrough
TreeView+ depends on / blocked
Reported: 2015-09-10 23:47 UTC by David Gibson
Modified: 2016-02-18 15:47 UTC (History)
14 users (show)

Fixed In Version: qemu-kvm-rhev-2.3.0-24.el7
Doc Type: Bug Fix
Doc Text:
Clone Of:
Last Closed: 2015-12-04 16:57:01 UTC
Target Upstream Version:

Attachments (Terms of Use)

System ID Private Priority Status Summary Last Updated
IBM Linux Technology Center 133677 0 None None None Never
Red Hat Product Errata RHBA-2015:2546 0 normal SHIPPED_LIVE qemu-kvm-rhev bug fix and enhancement update 2015-12-04 21:11:56 UTC

Description David Gibson 2015-09-10 23:47:50 UTC
Description of problem:

If qemu is configured with large amounts of hotpluggable memory (say > 256G) it takes a very long time to start the guest.  

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1. Start qemu with
      /usr/libexec/qemu-kvm -M pseries -m 512M,maxmem=XXXXX ...
For maxmem=512G, 1024G, 2048G
2. Time how long until qemu starts booting the guest

Actual results:

512G: ~17s
1024G: ~2m30s
2048G: ~20m

Expected results:

<10s startup in all cases.

Additional info:

The root cause is that qemu constructs a DR connector object internally for every 256M of hotpluggable memory.  The current algorithm for doing so is O(n^3).

Comment 1 David Gibson 2015-09-10 23:51:52 UTC
I've posted a patch upstream which improves adding the connectors to O(n^2).  This brings startup with 2048G to an acceptable 4s, unfortunately 4096G (RHEV's usual limit) is still ~35s.

Comment 4 David Gibson 2015-09-15 00:41:55 UTC
I have a patch for this in the works, but I'll need a bunch of ack flags ASAP if I'm to get it in for the next snapshot.

Comment 5 David Gibson 2015-09-15 01:12:54 UTC
I've made  draft brew build with the partial fix for slow startup at:


Comment 8 Qunfang Zhang 2015-09-15 07:07:39 UTC
(In reply to David Gibson from comment #5)
> I've made  draft brew build with the partial fix for slow startup at:
> https://brewweb.devel.redhat.com/taskinfo?taskID=9835517

Test with the above build, boot guest with the following configuration:

256G: ~ 10s
512G: ~10s
1024G: ~15S
2048G: After 5s, hit the issue in bug 1263039 comment 4.

And this bug could be reproduced on the official build qemu-kvm-rhev-2.3.0-22.el7.ppc64le:

256G: ~ 10s
512G: ~17s
1024G: 2~3 mins
2048G: after about 27 mins, hit bug 1263039.

Anyway, this qemu-kvm build fixes the original issue in this bug.

Comment 9 Miroslav Rezanina 2015-09-18 11:54:29 UTC
Fix included in qemu-kvm-rhev-2.3.0-24.el7

Comment 10 Qunfang Zhang 2015-09-24 05:23:32 UTC
Verified this bug with the following version:


Test with the same command line with comment 2 and the result is:

256G: ~ 10s
512G: ~10s without guest desktop (with guest desktop, it consumes 20+ seconds)
1024G: ~10s without guest desktop (with guest desktop, it consumes 20+ seconds)

According to bug 1263039 comment6, we limit the maxmem to 1T.

So according to comment 8 and the above result, I will set the status to VERIFIED.

Comment 11 IBM Bug Proxy 2015-11-30 15:20:36 UTC
------- Comment From fnovak.com 2015-11-30 15:17 EDT-------
reverse mirror of RHBZ 1262143 - VM startup is very slow with large amounts of hotpluggable memory

Comment 13 errata-xmlrpc 2015-12-04 16:57:01 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.