From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from firstgate.proxmox.com (firstgate.proxmox.com [212.224.123.68]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by lists.proxmox.com (Postfix) with ESMTPS id 01D2460FFE for ; Tue, 20 Oct 2020 08:25:44 +0200 (CEST) Received: from firstgate.proxmox.com (localhost [127.0.0.1]) by firstgate.proxmox.com (Proxmox) with ESMTP id E50F4ABBF for ; Tue, 20 Oct 2020 08:25:13 +0200 (CEST) Received: from proxmox-new.maurer-it.com (proxmox-new.maurer-it.com [212.186.127.180]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by firstgate.proxmox.com (Proxmox) with ESMTPS id 212ABABB2 for ; Tue, 20 Oct 2020 08:25:13 +0200 (CEST) Received: from proxmox-new.maurer-it.com (localhost.localdomain [127.0.0.1]) by proxmox-new.maurer-it.com (Proxmox) with ESMTP id DDABC45E2D for ; Tue, 20 Oct 2020 08:25:12 +0200 (CEST) To: Thomas Lamprecht , Proxmox VE development discussion References: <20201019105300.17898-1-f.ebner@proxmox.com> <7a6f8e5a-972d-166c-dafc-ccdd049226f7@proxmox.com> From: Fabian Ebner Message-ID: Date: Tue, 20 Oct 2020 08:25:07 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.3.1 MIME-Version: 1.0 In-Reply-To: <7a6f8e5a-972d-166c-dafc-ccdd049226f7@proxmox.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit X-SPAM-LEVEL: Spam detection results: 0 AWL -0.040 Adjusted score from AWL reputation of From: address KAM_DMARC_STATUS 0.01 Test Rule for DKIM or SPF Failure with Strict Alignment NICE_REPLY_A -0.001 Looks like a legit reply (A) RCVD_IN_DNSWL_MED -2.3 Sender listed at https://www.dnswl.org/, medium trust SPF_HELO_NONE 0.001 SPF: HELO does not publish an SPF Record SPF_PASS -0.001 SPF: sender matches SPF record URIBL_BLOCKED 0.001 ADMINISTRATOR NOTICE: The query to URIBL was blocked. See http://wiki.apache.org/spamassassin/DnsBlocklists#dnsbl-block for more information. [vzdump.pm, proxmox.com] Subject: Re: [pve-devel] [PATCH manager] get_included_guests: handle non-existing guests X-BeenThere: pve-devel@lists.proxmox.com X-Mailman-Version: 2.1.29 Precedence: list List-Id: Proxmox VE development discussion List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 20 Oct 2020 06:25:44 -0000 On 10/19/20 2:38 PM, Thomas Lamprecht wrote: > (FYI: forgot to hit reply-all, so resending this for the list) > > On 19.10.20 12:53, Fabian Ebner wrote: >> If a guest is removed without purge, the ID will remain >> in the backup configuration. Avoid using the variable $node >> when it is potentially undefined. Instead, skip non-existing >> guests and warn the user. >> >> Reported here: >> https://forum.proxmox.com/threads/purge-backup-does-not-remove-vm-from-datacenter-backup-list.77609/ > > a backup job referencing to an non-existent VM must fail, that's by design. Well, it currently doesn't; in fact no task is being run for such guests (123 doesn't exists, 124 is on a different node in this example): root@rob1 ~ # vzdump 121 123 124 --storage myfs --mode snapshot --remove 1 Use of uninitialized value $node in hash element at /usr/share/perl5/PVE/VZDump.pm line 1221. INFO: starting new backup job: vzdump 121 123 124 --storage myfs --mode snapshot --remove 1 INFO: skip external VMs: 124, 123 INFO: Starting Backup of VM 121 (qemu) ---8<--- The reason is that a missing guest ID is assigned to the $vmids_per_node hash with key equal to undef, which causes the hash to look like: $VAR1 = { '' => [ 123 ], 'rob2' => [ 124 ], 'rob1' => [ 121 ] }; and then back in the vzdump API call, it lands in the $skiplist. So there is no backup task run for missing guest IDs. > People should either use purge, if they really want to purge a VM, or else > remove it manually from the job. > > It's just important that the backup of the remaining, existing, VMs is made > nonetheless, but the job is not successful, as it was asked to backup > something that does not exists - making such errors less prominent is not > ideal. > With the patch we at least get a warning with the real problem and avoid using an undefined hash key. But sure, I'll work out a v2 where there is a failing backup task for missing guests. >> >> Signed-off-by: Fabian Ebner >> --- >> PVE/VZDump.pm | 5 +++++ >> test/vzdump_guest_included_test.pl | 12 +++++++++++- >> 2 files changed, 16 insertions(+), 1 deletion(-) >> >> diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm >> index 542228d6..6dbb6a44 100644 >> --- a/PVE/VZDump.pm >> +++ b/PVE/VZDump.pm >> @@ -1217,6 +1217,11 @@ sub get_included_guests { >> $vmids = check_vmids(@$vmids); >> >> for my $vmid (@$vmids) { >> + if (!defined($vmlist->{ids}->{$vmid})) { >> + debugmsg('warn', "no guest with ID '$vmid' exists in the cluster!"); >> + next; >> + } >> + >> my $node = $vmlist->{ids}->{$vmid}->{node}; >> next if (defined $job->{node} && $job->{node} ne $node); >> >