Gilles (Gilles Dubuc)
Senior Performance Engineer, WMF

Projects (20)

Today

  • Clear sailing ahead.

Tomorrow

  • Clear sailing ahead.

Sunday

  • Clear sailing ahead.

User Details

User Since
Oct 7 2014, 8:24 AM (158 w, 3 d)
Availability
Available
IRC Nick
gilles
LDAP User
Gilles
MediaWiki User
Unknown

Recent Activity

Yesterday

Gilles added a comment to T178559: Chrome 62 performance regression.

What would be the harm in switching to "defer" for both the startup module and the preloaded modules?

Thu, Oct 19, 7:38 PM · Performance-Team, WebPageTest
Gilles added a comment to T178559: Chrome 62 performance regression.

Right, it's unclear to what extent real users would be affected, but since we have no way of measuring visual completion with RUM, we can only assume it might have such a dramatic effect for some. I'd feel more comfortable if we could get back to the performance we used to have in synthetic tests. This stuff is all low priority, it should get out of the way during the critical path. Preload is still worthwhile if we can make it free of side effects.

Thu, Oct 19, 1:07 PM · Performance-Team, WebPageTest
Gilles added a comment to T178559: Chrome 62 performance regression.

Running via WPT? Or testing manually?

Thu, Oct 19, 12:43 PM · Performance-Team, WebPageTest
Gilles added a comment to T178559: Chrome 62 performance regression.

To summarize:

Thu, Oct 19, 12:38 PM · Performance-Team, WebPageTest
Gilles added a comment to T178559: Chrome 62 performance regression.

I assume that the pink bars ween in WPT are when the JS in the corresponding request is given execution/CPU time.

Thu, Oct 19, 12:23 PM · Performance-Team, WebPageTest
Gilles added a comment to T178559: Chrome 62 performance regression.

Last Chrome 61 unauthed desktop run: http://wpt.wmftest.org/result/171018_Z4_FE/1/details/
First Chrome 62 unauthed desktop run: http://wpt.wmftest.org/result/171018_PJ_G0/1/details/

Thu, Oct 19, 12:17 PM · Performance-Team, WebPageTest
Gilles updated subscribers of T175093: Decommission osmium.eqiad.wmnet.

Is it too late for us to run something on that machine before it's fully decommissioned? We've been trying something in labs for T176361 and we're at the point where we're wondering if bare metal would be better for what we're trying to achieve.

Thu, Oct 19, 11:56 AM · Patch-For-Review, Performance-Team (Radar), ops-eqiad, hardware-requests, Operations
Gilles renamed T178559: Chrome 62 performance regression from Investigate increase in render and SpeedIndex for Chrome to Chrome 62 performance regression.
Thu, Oct 19, 11:53 AM · Performance-Team, WebPageTest
Gilles added a comment to T178559: Chrome 62 performance regression.

Ok, so the difference is clearly Chrome 61 vs Chrome 62 behavior. I think it's safe to assume that nothing changed in our code.

Thu, Oct 19, 11:52 AM · Performance-Team, WebPageTest
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Are WPT and WebPageReplay at 10fps in the above breakdown? Given the ranges we're looking at, I think the rounding that happens at 10fps might make anything running at 10fps look more stable than it really is. A span of up to 200ms could be reduced to 100ms. Up to 300ms could end up being around 200ms, etc. And obviously it introduces a floor, where it's impossible to go below 100ms span at 10fps, even if the setup performs better.

Thu, Oct 19, 7:34 AM · Performance-Team

Wed, Oct 18

Gilles moved T177058: Set up performance tooling & dashboards from Inbox to Radar on the Performance-Team board.
Wed, Oct 18, 7:15 PM · Performance-Team (Radar), Epic, Marvin
Gilles moved T178072: Thumbor: Error reading image metadata: Failed to read image data from Inbox to Doing on the Performance-Team board.
Wed, Oct 18, 7:14 PM · Performance-Team, Thumbor
Gilles moved T178350: Provide better backend instrumentation for ResourceLoader from Inbox to Doing on the Performance-Team board.
Wed, Oct 18, 7:14 PM · MW-1.31-release-notes (WMF-deploy-2017-10-24 (1.31.0-wmf.5)), Patch-For-Review, Performance-Team, MediaWiki-ResourceLoader
Gilles moved T178352: Test measuring PWAs with lastVisualChange from Inbox to Next-up on the Performance-Team board.
Wed, Oct 18, 7:11 PM · Performance-Team, Marvin
Gilles moved T178479: Investigate navtiming2 values from Inbox to Next-up on the Performance-Team board.
Wed, Oct 18, 6:55 PM · Performance-Team
Gilles added a comment to T178457: nutcracker fails to start due to lack of /var/run/nutcracker (ex: deployment-videoscaler01 has memcached failures).

Videoscalers don't run Thumbor

Wed, Oct 18, 12:28 PM · Patch-For-Review, Operations, Release-Engineering-Team (Kanban), Beta-Cluster-Infrastructure
Gilles removed a project from T178457: nutcracker fails to start due to lack of /var/run/nutcracker (ex: deployment-videoscaler01 has memcached failures): Thumbor.
Wed, Oct 18, 12:28 PM · Patch-For-Review, Operations, Release-Engineering-Team (Kanban), Beta-Cluster-Infrastructure

Tue, Oct 17

Gilles created T178360: Consider having a light source casting shadows in the 3D viewer.
Tue, Oct 17, 9:55 AM · Multimedia, 3D
Gilles removed a project from T169522: Measure separate NavigationTiming metric(s) focused on Asia with higher sampling: Patch-For-Review.
Tue, Oct 17, 9:30 AM · MW-1.31-release-notes (WMF-deploy-2017-10-17 (1.31.0-wmf.4)), Performance-Team

Mon, Oct 16

Gilles updated subscribers of T176361: Run performance tests on VPS/Labs using local proxy.

We could possibly take an image scaler our of rotation on codfw, since those serve very few requests since the Thumbor deployment (and none in codfw outside of switchovers). @fgiunchedi is that something that sounds reasonable? If initial testing is promising, we could subsequently request a bare metal machine properly. Or talk about repurposing an image scaler for good, I guess.

Mon, Oct 16, 6:40 PM · Performance-Team
Gilles added a comment to T177073: Split the backend savetiming metric into submetrics.

Do we have dashboard for these new metrics yet?

Mon, Oct 16, 9:45 AM · MW-1.31-release-notes (WMF-deploy-2017-10-10 (1.31.0-wmf.3)), Patch-For-Review, Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

OK, I think the experiment of running browsertime alone has run its course, clearly webpagereplay is giving us greater stability on every metric. I'm going to destroy the browsertime machine and remove the corresponding labs dashboard.

Mon, Oct 16, 8:22 AM · Performance-Team
Gilles closed T177283: Get up to speed on Edge dev/perf tools as Resolved.

Meeting scheduled with the Edge team this week

Mon, Oct 16, 8:17 AM · Performance-Team
Gilles closed T175689: Remove X-Content-Dimensions for multipage originals as Resolved.

Cleanup should be complete for TIFFs, PDFs and DJVUs on all wikis.

Mon, Oct 16, 8:15 AM · MW-1.31-release-notes (WMF-deploy-2017-09-26 (1.31.0-wmf.1)), User-fgiunchedi, Operations, Performance-Team, Thumbor
Gilles closed T175689: Remove X-Content-Dimensions for multipage originals, a subtask of T167034: Limit maximum x-content-dimension size to avoid hitting nginx limits, as Resolved.
Mon, Oct 16, 8:15 AM · MW-1.30-release-notes, User-fgiunchedi, Patch-For-Review, Operations, Performance-Team, Thumbor

Fri, Oct 13

Gilles added a comment to T170628: Jessie rsvg/cairo can't render specific SVG file on Commons.

Probably something in the Cairo libraries? That's where you located the issue last time.

Fri, Oct 13, 7:43 PM · Performance-Team (Radar), Operations, Thumbor, Commons, media-storage
Elitre awarded T171093: Performance review of VisualEditor a Love token.
Fri, Oct 13, 12:14 PM · VisualEditor-Performance, VisualEditor, Performance-Team
Gilles committed rTHMBREXT2b1b705feea1: Avoid log noise coming from pyexiv2 (authored by Gilles).
Avoid log noise coming from pyexiv2
Fri, Oct 13, 8:37 AM
Gilles added a commit to T178072: Thumbor: Error reading image metadata: Failed to read image data: rTHMBREXT2b1b705feea1: Avoid log noise coming from pyexiv2.
Fri, Oct 13, 8:37 AM · Performance-Team, Thumbor
Gilles changed the status of T178072: Thumbor: Error reading image metadata: Failed to read image data from Open to Stalled.

I'm going to wait until there's a more compelling reason to push a new version. As it stands those noisy messages can easily be filtered out in Logstash.

Fri, Oct 13, 8:37 AM · Performance-Team, Thumbor
Gilles triaged T178072: Thumbor: Error reading image metadata: Failed to read image data as Low priority.
Fri, Oct 13, 8:33 AM · Performance-Team, Thumbor
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

throttle --up 330 --down 780 --rtt 200 is now running in a screen under my user on the browsertime machine, and the browsertime command on the cron job no longer has any connectivity parameter.

Fri, Oct 13, 6:31 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I will make the throttle permanent and see what happens

Fri, Oct 13, 6:27 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Just checked the latest har file, the slow runs are random.

Fri, Oct 13, 6:26 AM · Performance-Team
Gilles moved T177283: Get up to speed on Edge dev/perf tools from Next-up to Doing on the Performance-Team board.
Fri, Oct 13, 6:15 AM · Performance-Team
Gilles added a comment to T176211: Page Previews could load less JS on pageload.

We're talking about a 6.3% increase in SpeedIndex, that's nothing to sneeze at. Which these arguments it takes 3 teams who each consider their feature to be a special snowflake to get close to a 20% regression. The problem isn't the amount per se, but the fact that we can do something about it. In logical terms the bulk of that module doesn't need to be loaded and executed at that point in time.

Fri, Oct 13, 5:03 AM · Patch-For-Review, Performance-Team (Radar), Readers-Web-Backlog, Page-Previews
Gilles added a comment to T176262: Consider adding mw.loader.preload / OutputPage::preloadModules().

I think the parse/eval should also be done lazily. Ideally we have the data in a "black box", we know it's the contents of the module (we should know that because we know which modules we've requested), and we do all the main thread work for it lazily. This has the advantage of avoiding the parse/eval pile-up in the local storage case.

Fri, Oct 13, 4:58 AM · Performance-Team, MediaWiki-ResourceLoader
Gilles added a comment to T150734: Make Thumbor logs available in ELK.

The logs are flowing, but the host information is missing. It needs to be merged and deployed.

Fri, Oct 13, 4:22 AM · Patch-For-Review, User-fgiunchedi, Performance-Team, Thumbor

Thu, Oct 12

Gilles added a comment to T170628: Jessie rsvg/cairo can't render specific SVG file on Commons.

It's specifically resizing it takes issue with.

Thu, Oct 12, 3:13 PM · Performance-Team (Radar), Operations, Thumbor, Commons, media-storage
Gilles reopened T170628: Jessie rsvg/cairo can't render specific SVG file on Commons as "Open".

The upgrade hasn't fixed the problem. It seems like converting the file without options works, but not with the options we typically use:

Thu, Oct 12, 2:12 PM · Performance-Team (Radar), Operations, Thumbor, Commons, media-storage
Gilles reopened T170628: Jessie rsvg/cairo can't render specific SVG file on Commons, a subtask of T43371: Thumbnail/imagescaler (tracking), as Open.
Thu, Oct 12, 2:12 PM · media-storage, Commons, Tracking
Gilles added a comment to T178072: Thumbor: Error reading image metadata: Failed to read image data.

Seems like just noise coming from the pyexiv2 library, writing to the error log. The actual exception is caught by the code.

Thu, Oct 12, 1:55 PM · Performance-Team, Thumbor
Gilles added a comment to T178072: Thumbor: Error reading image metadata: Failed to read image data.

Yes, they're coming from exiv2. I need to check if they're merely spewing into the log or if they're really causing requests to fail.

Thu, Oct 12, 1:45 PM · Performance-Team, Thumbor
Gilles updated the task description for T178072: Thumbor: Error reading image metadata: Failed to read image data.
Thu, Oct 12, 1:44 PM · Performance-Team, Thumbor
Gilles created T178072: Thumbor: Error reading image metadata: Failed to read image data.
Thu, Oct 12, 1:38 PM · Performance-Team, Thumbor
Gilles closed T150734: Make Thumbor logs available in ELK as Resolved.

Success: https://logstash.wikimedia.org/goto/c016a93d8c2cedd96d4cd0937e7230dc

Thu, Oct 12, 1:25 PM · Patch-For-Review, User-fgiunchedi, Performance-Team, Thumbor
Gilles created T178043: WPT: Mobile Facebook article test often has cut off viewport.
Thu, Oct 12, 6:14 AM · Upstream, WebPageTest, Performance-Team
Gilles added a comment to T176211: Page Previews could load less JS on pageload.

Did you bother to look at this link I provided in my last comment?
https://phabricator.wikimedia.org/T175916#3608031 How can you be "unaware of a regression related to the launch" after looking at this?

Thu, Oct 12, 6:07 AM · Patch-For-Review, Performance-Team (Radar), Readers-Web-Backlog, Page-Previews
Gilles added a comment to T150734: Make Thumbor logs available in ELK.

@godog it's logstash that needs to be restarted

Thu, Oct 12, 5:58 AM · Patch-For-Review, User-fgiunchedi, Performance-Team, Thumbor

Wed, Oct 11

Gilles moved T177778: Improve database application performance monitoring visibility from Inbox to Radar on the Performance-Team board.
Wed, Oct 11, 5:27 PM · Performance-Team (Radar), MediaWiki-Platform-Team, monitoring, DBA
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

--connectivity.engine throttle -c cable gave me an error the first time I used it:

Wed, Oct 11, 11:49 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

--cacheClearRaw is undocumented on the front page of https://github.com/sitespeedio/browsertime btw

Wed, Oct 11, 11:18 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Without doing anything special, the browsertime metrics are more volatile. The min-max span for SpeedIndex is 5% of the min value on the webpagereplay server, 11% on the browsertime server.

Wed, Oct 11, 11:02 AM · Performance-Team
Gilles added a comment to T177747: grafana-labs often fails to generate graphs with c.datapoints is undefined.

Nice find!

Wed, Oct 11, 10:58 AM · Graphite, Cloud-VPS

Tue, Oct 10

Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

The browsertime machine (without any proxying) is set up and running, visible in Grafana. Since the runs are much faster, I have it running every 3 minutes to collect more data.

Tue, Oct 10, 11:53 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

It's now putting the run data into a tmpfs partition. I couldn't put the chrome data in there as well, because doing so reuses the same profile and cache between runs inside a given browsertime call. This is what was happening in the latest batch of mahimahi and could explain why it got better, it just started (accidentally) using the browser cache across runs.

Tue, Oct 10, 11:29 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I think the results we've got from Cloud VPS are ok. AWS is a bit better, but it's a lot more expensive and not on our infrastructure. When we started we didn't know what kind of stability we were going to get, and it seems to be as good as it gets. Which is already a lot better than what WPT without a proxy is giving us right now. Neither AWS not Cloud VPS can achieve commit-level stability yet, but already they're useful for better alerts. I don't think the difference in stability between the two justifies paying for AWS, it's still in the same ballpark.

Tue, Oct 10, 8:16 AM · Performance-Team

Mon, Oct 9

Gilles edited P6083 Set up mahimahi-h2o.
Mon, Oct 9, 3:59 PM · Performance-Team
Gilles edited P6081 Mahimahi run script.
Mon, Oct 9, 3:59 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I've tried pinning chromium to a different CPU than browsertime via a shell wrapper and it's not liking it, crashing randomly. I think that concludes my investigation of that idea. Pinning things to one CPU just makes stability and performance worse than letting the OS do its own scheduling.

Mon, Oct 9, 3:04 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Mon, Oct 9, 2:33 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I've set a CPU aside and giving it exclusively to browsertime and its subprocesses for the mahimahi machines. It's not ideal because ffmpeg, Xvfb and chrome are still sharing a CPU, but h2o and OS processes are definitely running on other CPUs. I don't know how useful that'll be if the physical CPUs are shared between VMs for Cloud VPS. Might be more useful for AWS, but there I see only 2 cores.

Mon, Oct 9, 2:15 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

For the mahimahi machine it definitely hasn't made things more stable than before (tmpfs for run data and chrome user data).

Mon, Oct 9, 1:37 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Mon, Oct 9, 1:22 PM · Performance-Team
Gilles edited P6081 Mahimahi run script.
Mon, Oct 9, 1:21 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I think tmpfs is safer because you reserve the space and it can't go over the limit

Mon, Oct 9, 1:12 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Mon, Oct 9, 1:03 PM · Performance-Team
Gilles added a comment to T177747: grafana-labs often fails to generate graphs with c.datapoints is undefined.

Yes, we run into the same problem occasionally in production: https://phabricator.wikimedia.org/T153169#3042185, but it's really painful on Labs Graphite/Grafana.

Mon, Oct 9, 12:45 PM · Graphite, Cloud-VPS
Gilles added a comment to T177747: grafana-labs often fails to generate graphs with c.datapoints is undefined.

It's Graphite intermittently serving junk to Grafana. It's a PNG image instead of JSON being sent back. I've seen that before, but I can't remember if it was production Graphite or that I'm just rediscovering the same issue on Labs Graphite again...

Mon, Oct 9, 12:36 PM · Graphite, Cloud-VPS
Gilles updated subscribers of T177747: grafana-labs often fails to generate graphs with c.datapoints is undefined.
Mon, Oct 9, 12:33 PM · Graphite, Cloud-VPS
Gilles edited P6083 Set up mahimahi-h2o.
Mon, Oct 9, 12:31 PM · Performance-Team
Gilles edited P6081 Mahimahi run script.
Mon, Oct 9, 12:30 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Mon, Oct 9, 12:27 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I've just noticed an interesting difference: on AWS the underlying disk is an SSD, on Cloud VPS it's a spinning disk. We might get better perf by having the run read and write its data to a RAM partition.

Mon, Oct 9, 12:13 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

...and that's already what you're doing :) OK, it doesn't seem like there's a better option with ffmpeg. Which doesn't mean that we couldn't benefit from isolating things per CPU more. I think it's a matter of figuring out what consumes the most CPU during runs (eg. could be the web servers replaying stuff).

Mon, Oct 9, 12:09 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

But ideally raw video would be best (no compression at all). I've just seen people sharing their screen recording ffmpeg code and they use this: "-vcodec libx264 -vpre lossless_ultrafast" which suggests that it will be contained in a h.264 format but compressed as little as possible.

Mon, Oct 9, 12:07 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Reading your commit, it seems like you've decoupled recording and encoding. It's just a matter of changing the codec used during recording to be one with the lowest overhead possible.

Mon, Oct 9, 12:03 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Or is what I'm describing already what browsertime doing?

Mon, Oct 9, 11:56 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

You can get the best of both worlds and record the video uncompressed during the run, then once you're done measuring, encode it to x264 or whatever you like. This way the huge raw video file doesn't stick around, it's just kept for as long as you need to derive metrics from it. You could even encode before measuring metrics, it really doesn't matter as long as perf measurements aren't happening at the same time as expensive video encoding.

Mon, Oct 9, 11:54 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

x264 is going to be expensive to encode regardless. Why are we compressing video at all? It's useful if you're going to expose the videos to be played on the web, but in terms of figuring out the SpeedIndex, etc. it's just overhead at the time the measurements are being made.

Mon, Oct 9, 11:50 AM · Performance-Team
Gilles added a comment to T177283: Get up to speed on Edge dev/perf tools.

I think I'm done touring the dev tools. My biggest pet peeve is the inability to contrast the performance timeline with screenshots.

Mon, Oct 9, 11:47 AM · Performance-Team
Gilles added a comment to T177283: Get up to speed on Edge dev/perf tools.

https://developer.microsoft.com/en-us/microsoft-edge/tools/vms/ has Edge 16 VMs

Mon, Oct 9, 9:09 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Sigh, actually you need to nuke the entire machine, since the .ssh folder is gone, can't ssh into it anymore.

Mon, Oct 9, 8:35 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

I've accidentally nuked the home directory of the mahimahi AWS machine, can you give me the IP address of the host to send graphite metrics to?

Mon, Oct 9, 8:32 AM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

It's certainly an improvement compared to https://grafana.wikimedia.org/dashboard/db/webpagetest-drilldown?orgId=1&var-wiki=enwiki&var-users=anonymous&var-page=Barack_Obama&var-location=us-east-1&var-browser=Google_Chrome&var-view=firstView but I'm disappointed in the "best" results on AWS. That means we can't catch regressions of less than 5% reliably. Maybe we should experiment with reserving CPUs for certain tasks. For example booting the kernel with N-1 CPUs available, and ffmpeg runs exclusively on the Nth CPU?

Mon, Oct 9, 7:26 AM · Performance-Team

Fri, Oct 6

Gilles added a comment to T175689: Remove X-Content-Dimensions for multipage originals.

Cleanup of non-Commons wikis complete, started the cleanup of Commons. There are 1.2 million files to process there, at the current rate it's going to take approximately 9 days to complete.

Fri, Oct 6, 7:55 AM · MW-1.31-release-notes (WMF-deploy-2017-09-26 (1.31.0-wmf.1)), User-fgiunchedi, Operations, Performance-Team, Thumbor

Thu, Oct 5

Gilles merged task T168526: Set up a server running mitmproxy + mahimahi to measure key synthetic testing over time into T176361: Run performance tests on VPS/Labs using local proxy.
Thu, Oct 5, 1:30 PM · Performance-Team
Gilles merged T168526: Set up a server running mitmproxy + mahimahi to measure key synthetic testing over time into T176361: Run performance tests on VPS/Labs using local proxy.
Thu, Oct 5, 1:30 PM · Performance-Team
Gilles added a comment to T176361: Run performance tests on VPS/Labs using local proxy.

Command to set up a fresh Debian 9 machine with mahimahi, with the custom script to send metrics to graphite:

Thu, Oct 5, 1:30 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Thu, Oct 5, 1:07 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Thu, Oct 5, 1:05 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Thu, Oct 5, 12:58 PM · Performance-Team
Gilles edited P6083 Set up mahimahi-h2o.
Thu, Oct 5, 12:54 PM · Performance-Team
Gilles created P6083 Set up mahimahi-h2o.
Thu, Oct 5, 12:38 PM · Performance-Team
Gilles created P6081 Mahimahi run script.
Thu, Oct 5, 11:50 AM · Performance-Team
Gilles added a comment to T175689: Remove X-Content-Dimensions for multipage originals.

Started running the cleanup script on Terbium, clearing the header for TIFFs, PDFs and DJVUs.

Thu, Oct 5, 9:50 AM · MW-1.31-release-notes (WMF-deploy-2017-09-26 (1.31.0-wmf.1)), User-fgiunchedi, Operations, Performance-Team, Thumbor

Wed, Oct 4

Gilles added a comment to T169522: Measure separate NavigationTiming metric(s) focused on Asia with higher sampling.

Geo.region doesn't contain the continent.

Wed, Oct 4, 7:41 PM · MW-1.31-release-notes (WMF-deploy-2017-10-17 (1.31.0-wmf.4)), Performance-Team
Gilles removed a project from T175689: Remove X-Content-Dimensions for multipage originals: Patch-For-Review.
Wed, Oct 4, 7:36 PM · MW-1.31-release-notes (WMF-deploy-2017-09-26 (1.31.0-wmf.1)), User-fgiunchedi, Operations, Performance-Team, Thumbor
Gilles removed a project from T169522: Measure separate NavigationTiming metric(s) focused on Asia with higher sampling: Patch-For-Review.
Wed, Oct 4, 7:35 PM · MW-1.31-release-notes (WMF-deploy-2017-10-17 (1.31.0-wmf.4)), Performance-Team
Gilles moved T175238: Write extensive on-wiki documentation about Thumbor from Next-up to Doing on the Performance-Team board.
Wed, Oct 4, 7:29 PM · Performance-Team, Thumbor