public inbox for pve-devel@lists.proxmox.com
 help / color / mirror / Atom feed
From: "Fabian Grünbichler" <f.gruenbichler@proxmox.com>
To: Fabian Ebner <f.ebner@proxmox.com>, pve-devel@lists.proxmox.com
Subject: Re: [pve-devel] [PATCH http-server 1/1] webproxy: handle unflushed write buffer
Date: Mon, 08 Nov 2021 16:45:19 +0100	[thread overview]
Message-ID: <1636385624.81iw4jegre.astroid@nora.none> (raw)
In-Reply-To: <9df61960-edb3-1dc2-5e75-993817e9f024@proxmox.com>

On November 8, 2021 3:15 pm, Fabian Ebner wrote:
> Am 05.11.21 um 14:03 schrieb Fabian Grünbichler:
>> for proxied requests, we usually tear down the proxy connection
>> immediately when closing the source connection. this is not the correct
>> course of action for bulk one-way data streams that are proxied, where
>> the source connection might be closed, but the proxy connection might
>> still have data in the write buffer that needs to be written out.
>> 
>> push_shutdown already handles this case (closing the socket/FH after it
>> has been fully drained).
>> 
>> one example for such a proxied data stream is the 'migrate' data for a
>> remote migration, which gets proxied over a websocket connection.
>> terminating the proxied connection early makes the target VM crash for
>> obvious reasons.
>> 
>> Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
>> ---
>>   src/PVE/APIServer/AnyEvent.pm | 10 ++++++----
>>   1 file changed, 6 insertions(+), 4 deletions(-)
>> 
>> diff --git a/src/PVE/APIServer/AnyEvent.pm b/src/PVE/APIServer/AnyEvent.pm
>> index 86d0e2e..ecf771f 100644
>> --- a/src/PVE/APIServer/AnyEvent.pm
>> +++ b/src/PVE/APIServer/AnyEvent.pm
>> @@ -144,7 +144,8 @@ sub client_do_disconnect {
>>       };
>>   
>>       if (my $proxyhdl = delete $reqstate->{proxyhdl}) {
>> -	&$shutdown_hdl($proxyhdl);
>> +	&$shutdown_hdl($proxyhdl)
>> +		if !$proxyhdl->{block_disconnect};
> 
> Style nit: fits in one line ;)
> 
> I'm not familiar with the code, so I'll just ask: can this be reached 
> without going through the code below first, i.e. can it happen that 
> block_disconnect is not set, but length $proxyhdl->{wbuf} > 0? Or is it 
> not important in other cases (if there are any)?

in theory, yes. in practice, not likely with anything that matters ;)

we have

- spiceproxy, if the client closes the connection we likely 
  don't care about the last few bytes of input being dropped
- proxied API requests with response_stream - never gets written to
- WS proxy (fixed in this patch ;) technically also used like 
  spiceproxy)

but I have to admit I did think about doing this always and relying on 
timeouts to clear the proxyhdl eventually, but then decided to play it 
safe and just change the one I am sure requires it..

> 
>>       }
>>   
>>       my $hdl = delete $reqstate->{hdl};
>> @@ -627,9 +628,10 @@ sub websocket_proxy {
>>   		    } elsif ($opcode == 8) {
>>   			my $statuscode = unpack ("n", $payload);
>>   			$self->dprint("websocket received close. status code: '$statuscode'");
>> -			if ($reqstate->{proxyhdl}) {
>> -			    $reqstate->{proxyhdl}->push_shutdown();
>> -			}
>> +			if (my $proxyhdl = $reqstate->{proxyhdl}) {
>> +			    $proxyhdl->{block_disconnect} = 1 if length $proxyhdl->{wbuf} > 0;
>> +			    $proxyhdl->push_shutdown();
>> +		        }
>>   			$hdl->push_shutdown();
>>   		    } elsif ($opcode == 9) {
>>   			# ping received, schedule pong
>> 
> 




  reply	other threads:[~2021-11-08 15:46 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-05 13:03 [pve-devel] [PATCH-series qemu-server++ 0/22] remote migration Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH proxmox 1/1] websocket: adapt for client connection Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH proxmox-websocket-tunnel 1/4] initial commit Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH proxmox-websocket-tunnel 2/4] add tunnel implementation Fabian Grünbichler
2021-11-09 12:54   ` Dominik Csapak
2021-11-11  9:58     ` Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH proxmox-websocket-tunnel 3/4] add fingerprint validation Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH proxmox-websocket-tunnel 4/4] add packaging Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH access-control 1/2] tickets: add tunnel ticket Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH access-control 2/2] ticket: normalize path for verification Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH common 1/3] schema: rename storagepair to storage-pair Fabian Grünbichler
2021-11-11 13:18   ` [pve-devel] applied: " Thomas Lamprecht
2021-11-05 13:03 ` [pve-devel] [PATCH common 2/3] schema: add pve-bridge-id option/format/pair Fabian Grünbichler
2021-11-11 13:18   ` [pve-devel] applied: " Thomas Lamprecht
2021-11-05 13:03 ` [pve-devel] [PATCH common 3/3] schema: add proxmox-remote format/option Fabian Grünbichler
2021-11-11 13:18   ` [pve-devel] applied: " Thomas Lamprecht
2021-11-05 13:03 ` [pve-devel] [PATCH guest-common 1/1] migrate: handle migration_network with remote migration Fabian Grünbichler
2021-11-08 13:50   ` Fabian Ebner
2021-11-10 12:03   ` Fabian Ebner
2021-11-05 13:03 ` [pve-devel] [PATCH http-server 1/1] webproxy: handle unflushed write buffer Fabian Grünbichler
2021-11-08 14:15   ` Fabian Ebner
2021-11-08 15:45     ` Fabian Grünbichler [this message]
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 01/10] d/control: add pve-ha-manager to B-D Fabian Grünbichler
2021-11-11 13:18   ` [pve-devel] applied: " Thomas Lamprecht
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 02/10] adapt to renamed storage-pair format Fabian Grünbichler
2021-11-11 13:18   ` [pve-devel] applied: " Thomas Lamprecht
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 03/10] migrate: factor out storage checks Fabian Grünbichler
2021-11-11 13:18   ` [pve-devel] applied: " Thomas Lamprecht
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 04/10] refactor map_storage to map_id Fabian Grünbichler
2021-11-09  9:06   ` Fabian Ebner
2021-11-09 12:44     ` Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 05/10] schema: use pve-bridge-id Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 06/10] update_vm: allow simultaneous setting of boot-order and dev Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 07/10] mtunnel: add API endpoints Fabian Grünbichler
2021-11-09 12:46   ` Fabian Ebner
2021-11-10  7:40     ` Fabian Ebner
2021-11-11 11:07       ` Fabian Grünbichler
2021-11-11 11:04     ` Fabian Grünbichler
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 08/10] migrate: refactor remote VM/tunnel start Fabian Grünbichler
2021-11-09 14:04   ` Fabian Ebner
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 09/10] migrate: add remote migration handling Fabian Grünbichler
2021-11-10 11:17   ` Fabian Ebner
2021-11-11 12:25     ` Fabian Grünbichler
2021-11-11 12:57       ` Fabian Ebner
2021-11-05 13:03 ` [pve-devel] [PATCH qemu-server 10/10] api: add remote migrate endpoint Fabian Grünbichler
2021-11-10 12:29   ` Fabian Ebner
2021-11-11 12:33     ` Fabian Grünbichler
2021-11-09 16:07 ` [pve-devel] [PATCH-series qemu-server++ 0/22] remote migration DERUMIER, Alexandre
2021-11-10 12:42 ` Fabian Ebner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1636385624.81iw4jegre.astroid@nora.none \
    --to=f.gruenbichler@proxmox.com \
    --cc=f.ebner@proxmox.com \
    --cc=pve-devel@lists.proxmox.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Service provided by Proxmox Server Solutions GmbH | Privacy | Legal