[remoting host] "Probe delay too high" warning is spammy |
||||||
Issue descriptionWhen connecting via the website, the host logs many lines from bitrate_prober.cc: [0921/105127.811383:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297730187, now_ms: 1297730191) [0921/105127.816203:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297730189, now_ms: 1297730195) [0921/105132.865386:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297735241, now_ms: 1297735245) [0921/105132.869922:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297735243, now_ms: 1297735249) [0921/105132.875300:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297735245, now_ms: 1297735254) [0921/105132.880013:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297735247, now_ms: 1297735259) [0921/105138.154028:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297740529, now_ms: 1297740533) [0921/105138.158836:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297740531, now_ms: 1297740538) [0921/105138.163592:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297740533, now_ms: 1297740543) [0921/105138.168506:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297740535, now_ms: 1297740548) [0921/105142.974361:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297745350, now_ms: 1297745354) [0921/105142.979652:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297745352, now_ms: 1297745359) [0921/105148.184801:WARNING:bitrate_prober.cc(129)] Probe delay too high (next_ms:1297750560, now_ms: 1297750564)
,
Sep 24
It's even more spammy with a Firefox client, printing out hundreds of such messages every second.
,
Oct 2
Ping Sergey.
,
Oct 2
This message was added in https://webrtc-review.googlesource.com/c/src/+/30681 . Description on that CL lacks details - it's not clear what was the motivation behind it. The problem here is that thread wakes up too late to send next packet for the BW probe, which means that it's likely to provide incorrect results. Previously it would cause the probe to be restarted, which doesn't happen anymore.
,
Oct 3
I don't think this is important. I think philipel@ knows more about it.
,
Oct 17
philipel@, can you comment on this? We'd really like to remove this message, or find some way of disabling it for CRD.
,
Oct 18
When probing timing has to be fairly precise. Before we reset the probe attempt and tried again, but we removed that and instead added this warning to indicate that something interfered with the probe. I think you should investigate why you even get these warnings.
,
Oct 18
,
Oct 19
If you get this message hundreds of time per second then you should really investigate what is going on. Probing is at most done ~10 times per minute, and a typical probe cluster is 5-10 packets (for higher bitrates it can be more though), so even if there was a timing issue with every packet then you should still only see this message about 50~100 times per minute.
,
Oct 25
The following revision refers to this bug: https://webrtc.googlesource.com/src.git/+/9c8ae4b9a13cc87429d94181e7dd945dc9ae388b commit 9c8ae4b9a13cc87429d94181e7dd945dc9ae388b Author: Jamie Walch <jamiewalch@google.com> Date: Thu Oct 25 16:00:16 2018 Disable probe delay warning in release builds. This log is triggering many times a second for Chrome Remote Desktop on some browsers. This CL just turns it off for release builds to avoid log files filling up users' disks until we figure out what's going on. Bug: chromium:888038 Change-Id: Ibbe9d47295b3633314feb28e155e3f59b878dbdb Reviewed-on: https://webrtc-review.googlesource.com/c/107688 Commit-Queue: Jamie Walch <jamiewalch@google.com> Reviewed-by: Philip Eliasson <philipel@webrtc.org> Cr-Commit-Position: refs/heads/master@{#25372} [modify] https://crrev.com/9c8ae4b9a13cc87429d94181e7dd945dc9ae388b/modules/pacing/bitrate_prober.cc
,
Oct 25
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/d6003da075b745865943174c7d44a18f05097e32 commit d6003da075b745865943174c7d44a18f05097e32 Author: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com> Date: Thu Oct 25 23:09:22 2018 Roll src/third_party/webrtc 6c6c9df99db6..8c27ccac7597 (4 commits) https://webrtc.googlesource.com/src.git/+log/6c6c9df99db6..8c27ccac7597 git log 6c6c9df99db6..8c27ccac7597 --date=short --no-merges --format='%ad %ae %s' 2018-10-25 benwright@webrtc.org Promotoing webrtc::CryptoOptions to RTCConfiguration. 2018-10-25 benwright@webrtc.org Fixes use after free error when setting a new FrameEncryptor on ChannelSend. 2018-10-25 scottmg@chromium.org fuchsia: Stub out timing and memory functions 2018-10-25 jamiewalch@google.com Disable probe delay warning in release builds. Created with: gclient setdep -r src/third_party/webrtc@8c27ccac7597 The AutoRoll server is located here: https://autoroll.skia.org/r/webrtc-chromium-autoroll Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+/master/autoroll/README.md If the roll is causing failures, please contact the current sheriff, who should be CC'd on the roll, and stop the roller if necessary. CQ_INCLUDE_TRYBOTS=luci.chromium.try:linux_chromium_archive_rel_ng;luci.chromium.try:mac_chromium_archive_rel_ng BUG= chromium:808287 ,chromium:888038 TBR=webrtc-chromium-sheriffs-robots@google.com Change-Id: I6bda4329d38b26019758d06a637ffa44712f12c2 Reviewed-on: https://chromium-review.googlesource.com/c/1299590 Reviewed-by: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com> Commit-Queue: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com> Cr-Commit-Position: refs/heads/master@{#602918} [modify] https://crrev.com/d6003da075b745865943174c7d44a18f05097e32/DEPS
,
Jan 11
This issue has an owner, a component and a priority, but is still listed as untriaged or unconfirmed. By definition, this bug is triaged. Changing status to "assigned". Please reach out to me if you disagree with how I've done this. |
||||||
►
Sign in to add a comment |
||||||
Comment 1 by lambroslambrou@chromium.org
, Sep 24