From mboxrd@z Thu Jan 1 00:00:00 1970
Return-Path: <s.reiter@proxmox.com>
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 A18886BAB1
for <pve-devel@lists.proxmox.com>; Wed, 27 Jan 2021 12:11:07 +0100 (CET)
Received: from firstgate.proxmox.com (localhost [127.0.0.1])
by firstgate.proxmox.com (Proxmox) with ESMTP id 96E5B21AFE
for <pve-devel@lists.proxmox.com>; Wed, 27 Jan 2021 12:11:07 +0100 (CET)
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 2826A21AEE
for <pve-devel@lists.proxmox.com>; Wed, 27 Jan 2021 12:11:04 +0100 (CET)
Received: from proxmox-new.maurer-it.com (localhost.localdomain [127.0.0.1])
by proxmox-new.maurer-it.com (Proxmox) with ESMTP id DBD5C4611F
for <pve-devel@lists.proxmox.com>; Wed, 27 Jan 2021 12:11:03 +0100 (CET)
To: Thomas Lamprecht <t.lamprecht@proxmox.com>,
Proxmox VE development discussion <pve-devel@lists.proxmox.com>
References: <20210114153921.16739-1-s.reiter@proxmox.com>
<37a43b7e-1919-bc0b-ac84-08411c86bd4d@proxmox.com>
From: Stefan Reiter <s.reiter@proxmox.com>
Message-ID: <a8286e0c-fc75-3128-4abb-dac1d0d643c3@proxmox.com>
Date: Wed, 27 Jan 2021 12:11:01 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
Thunderbird/78.6.1
MIME-Version: 1.0
In-Reply-To: <37a43b7e-1919-bc0b-ac84-08411c86bd4d@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 -0.037 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. [restenvironment.pm]
Subject: Re: [pve-devel] [PATCH common] allow longer timeout for cancelling
'vzdump' jobs
X-BeenThere: pve-devel@lists.proxmox.com
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Proxmox VE development discussion <pve-devel.lists.proxmox.com>
List-Unsubscribe: <https://lists.proxmox.com/cgi-bin/mailman/options/pve-devel>,
<mailto:pve-devel-request@lists.proxmox.com?subject=unsubscribe>
List-Archive: <http://lists.proxmox.com/pipermail/pve-devel/>
List-Post: <mailto:pve-devel@lists.proxmox.com>
List-Help: <mailto:pve-devel-request@lists.proxmox.com?subject=help>
List-Subscribe: <https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel>,
<mailto:pve-devel-request@lists.proxmox.com?subject=subscribe>
X-List-Received-Date: Wed, 27 Jan 2021 11:11:07 -0000
On 26/01/2021 19:23, Thomas Lamprecht wrote:
> On 14.01.21 16:39, Stefan Reiter wrote:
>> This attempts to solve the issue where on slow network storages,
>> aborting a backup job (which may wait for buffers to flush) could take
>> longer than 5 seconds, and would thus result in the task being killed by
>> SIGKILL, not removing the backup lock in the process.
>>
>> Make the implementation future-proof by using a map from task type to a
>> timeout value. Default stays at 5, so tasks other than 'vzdump' are not
>> affected.
>>
>> Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
>> ---
>> src/PVE/RESTEnvironment.pm | 16 ++++++++++++----
>> 1 file changed, 12 insertions(+), 4 deletions(-)
>>
>
> Not to sure about that map there in pve-common, that module should stay rather
> agnostic of user special treatment.
>
> Did you thought about passing that explicitly on worker creation, or setting it
> in the RPCEnv inside a worker?
I generally agree that it's a bit misplaced, but I don't see a way to
encode it in the worker - the only info we have in check_worker and
stop_task is the UPID, and I don't think it makes sense to encode a
timeout in that? Or is there a way I'm not seeing to retrieve additional
info about a worker from the UPID alone?
We could at least put the map in pve-manager, but I'm not sure if that's
any better.
>
>> diff --git a/src/PVE/RESTEnvironment.pm b/src/PVE/RESTEnvironment.pm
>> index d5b84d0..8a0cb9a 100644
>> --- a/src/PVE/RESTEnvironment.pm
>> +++ b/src/PVE/RESTEnvironment.pm
>> @@ -365,8 +365,16 @@ sub active_workers {
>> return $res;
>> }
>>
>> +my $timeout_map = {
>> + # backup cancellation on slow target storages might take a while, avoid
>> + # leaving the VM in locked state
>> + "vzdump" => 60,
>> +};
>> +
>> my $kill_process_group = sub {
>> - my ($pid, $pstart) = @_;
>> + my ($pid, $pstart, $timeout) = @_;
>> +
>> + $timeout //= 5;
>>
>> # send kill to process group (negative pid)
>> my $kpid = -$pid;
>> @@ -374,8 +382,7 @@ my $kill_process_group = sub {
>> # always send signal to all pgrp members
>> kill(15, $kpid); # send TERM signal
>>
>> - # give max 5 seconds to shut down
>> - for (my $i = 0; $i < 5; $i++) {
>> + for (my $i = 0; $i < $timeout; $i++) {
>> return if !PVE::ProcFSTools::check_process_running($pid, $pstart);
>> sleep (1);
>> }
>> @@ -394,7 +401,8 @@ sub check_worker {
>> return 0 if !$running;
>>
>> if ($killit) {
>> - &$kill_process_group($task->{pid});
>> + my $type = $task->{type};
>> + &$kill_process_group($task->{pid}, undef, $timeout_map->{$type});
>> return 0;
>> }
>>
>>
>
>