Xoogler in the cloud here. I have a very low qps service that serves HTML plus the follow-up resources. So it typically sits idle and then receives something in the order of 20 requests over 5s with concurrency well below 10, where concurrency limit is 80. I observe that clients regularly receive 429s from Cloud Run, typically after periods of service inactivity, even though an instance is still up (so it's not a cold-start problem). This can either be on the first request but often somewhere in the middle of the sequence (i.e. icons, css don't load).
The instance is concurrent, responsive and could easily handle the load, but Cloud Run doesn't let it. No other instances are spun up either, although we're not even at the max of 2. This suggests that Cloud Run for some reason estimates >2 instances needed?
Here's a typical request sequence, redacted from the logs:
... 20 min idle ...
I 2020-03-27T18:21:27.619317Z GET 307 288 B 5 ms
I 2020-03-27T18:21:27.706580Z GET 302 0 B 0 ms
I 2020-03-27T18:21:27.760271Z GET 200 5.83 KiB 5 ms
I 2020-03-27T18:21:27.838066Z GET 200 1.89 KiB 4 ms
I 2020-03-27T18:21:27.882751Z GET 200 1.05 KiB 4 ms
I 2020-03-27T18:21:27.886743Z GET 200 582 B 3 ms
I 2020-03-27T18:21:27.893060Z GET 200 533 B 4 ms
I 2020-03-27T18:21:27.897352Z GET 200 5.35 KiB 4 ms
I 2020-03-27T18:21:27.899086Z GET 200 11.38 KiB 6 ms
I 2020-03-27T18:21:27.905967Z GET 200 22.48 KiB 13 ms
I 2020-03-27T18:21:27.906113Z GET 200 592 B 13 ms
I 2020-03-27T18:21:27.907967Z GET 200 35.08 KiB 14 ms
...500ms...
I 2020-03-27T18:21:28.434846Z GET 200 2.76 MiB 50 ms
I 2020-03-27T18:21:28.465552Z GET 200 2.29 MiB 67 ms <= up to here all resources served from image
...2500ms...
I 2020-03-27T18:21:31.086943Z GET 200 2.95 KiB 706 ms <= IO-bound, talking to backend api
...1600ms...
W 2020-03-27T18:21:32.674973Z GET 429 14 B 0 ms <= !!!
W 2020-03-27T18:21:32.675864Z GET 429 14 B 0 ms <= !!!
W 2020-03-27T18:21:32.676292Z GET 429 14 B 0 ms <= !!!
I 2020-03-27T18:21:32.684265Z GET 200 547 B 6 ms
I 2020-03-27T18:21:32.686695Z GET 200 504 B 9 ms
I 2020-03-27T18:21:32.690580Z GET 200 486 B 12 ms
Conceivably that last group of requests are 6 parallel requests. Why would three be denied and three served? The service is way under capacity. A couple of reloads typically solve the issue.
It really appears to me as if the algorithm vastly overestimates the required resources after a period of inactivity. I'm happy to try a larger max-instances (redeployed to 10 now) but something really seems off with the estimates on the low end of the spectrum. If "2" as a concurrency setting is below what the platform supports, gcloud probably should probably enforce a higher minimum in the first place.
This is somewhat sad as it impacts people just "trying out" Cloud Run and they observe intermittent errors (partially rendered pages, ...) - which are even pinned on the client (4xx) who is certainly not at fault.
Happy to provide more data.
Configuration:
template:
metadata:
...
annotations:
...
autoscaling.knative.dev/maxScale: '2'
spec:
timeoutSeconds: 900
...
containerConcurrency: 80
containers:
...
resources:
limits:
cpu: 1000m
memory: 244Mi
This looks like a known issue with Cloud Run, I would recommend starring it to receive notifications and expedite resolution.
Related
We have build Ejabberd in AWS EC2 instance and have enabled the clustering in the 6 Ejabberd servers in Tokyo, Frankfurt, and Singapore regions.
The OS, middleware, applications and settings for each EC2 instance are exactly the same.
But currently, the Ejabberd CPUs in the Frankfurt and Singapore regions are overloaded.
The CPU of Ejabberd in the Japan region is normal.
Could you please tell me the suspicious part?
You can take a look at the ejabberd log files of the problematic (and the good) nodes, maybe you find some clue.
You can use the undocumented "ejabberdctl etop" shell command in the problematic nodes. It's similar to "top", but runs inside the erlang virtual machine that runs ejabberd
ejabberdctl etop
========================================================================================
ejabberd#localhost 16:00:12
Load: cpu 0 Memory: total 44174 binary 1320
procs 277 processes 5667 code 20489
runq 1 atom 984 ets 5467
Pid Name or Initial Func Time Reds Memory MsgQ Current Function
----------------------------------------------------------------------------------------
<9135.1252.0> caps_requests_cache 2393 1 2816 0 gen_server:loop/7
<9135.932.0> mnesia_recover 480 39 2816 0 gen_server:loop/7
<9135.1118.0> dets:init/2 71 2 5944 0 dets:open_file_loop2
<9135.6.0> prim_file:start/0 63 1 2608 0 prim_file:helper_loo
<9135.1164.0> dets:init/2 56 2 4072 0 dets:open_file_loop2
<9135.818.0> disk_log:init/2 49 2 5984 0 disk_log:loop/1
<9135.1038.0> ejabberd_listener:in 31 2 2840 0 prim_inet:accept0/3
<9135.1213.0> dets:init/2 31 2 5944 0 dets:open_file_loop2
<9135.1255.0> dets:init/2 30 2 5944 0 dets:open_file_loop2
<9135.0.0> init 28 1 3912 0 init:loop/1
========================================================================================
I am unsure if this is the platform to ask. But hopefully it is :).
I've got a 3 node setup of ceph.
node1
mds.node1 , mgr.node1 , mon.node1 , osd.0 , osd.1 , osd.6
14.2.22
node2
mds.node2 , mon.node2 , osd.2 , osd.3 , osd.7
14.2.22
node3
mds.node3 , mon.node3 , osd.4 , osd.5 , osd.8
14.2.22
For some reason though, When I down one node, It does not start backfilling/recovery at all. It just reports 3 osd's down as below. But does nothing to repair it....
If I run a ceph -s I get the below ouput:
[root#node1 testdir]# ceph -s
cluster:
id: 8932b76b-282b-4385-bee8-5c295af88e74
health: HEALTH_WARN
3 osds down
1 host (3 osds) down
Degraded data redundancy: 30089/90267 objects degraded (33.333%), 200 pgs degraded, 512 pgs undersized
1/3 mons down, quorum node1,node2
services:
mon: 3 daemons, quorum node1,node2 (age 2m), out of quorum: node3
mgr: node1(active, since 48m)
mds: homeFS:1 {0=node1=up:active} 1 up:standby-replay
osd: 9 osds: 6 up (since 2m), 9 in (since 91m)
data:
pools: 4 pools, 512 pgs
objects: 30.09k objects, 144 MiB
usage: 14 GiB used, 346 GiB / 360 GiB avail
pgs: 30089/90267 objects degraded (33.333%)
312 active+undersized
200 active+undersized+degraded
io:
client: 852 B/s rd, 2 op/s rd, 0 op/s wr
[root#node1 testdir]#
The odd thing though, when I boot up my 3rd node again it does recover and sync. But it looks like it's backfilling just not starting at all...
Is there something that might be causing it?
Update
What I did notice, If I mark a drive as out, it does recover it... But when the server node's down, and the drive's marked as out, it then does not recover it at all...
Update 2:
I noticed while experimenting that if the OSD is up, but out, It does recover... When the OSD is marked as down it does not begin to recover at all...
The default is 10 minutes for ceph to wait until it marks OSDs as out (mon_osd_down_out_interval). This can help in case a server just needs a reboot and returns within 10 minutes then all is good. If you need a longer maintenance window but you're not sure if it will be longer than 10 minutes, but the server will eventually return, set ceph osd set noout to prevent unnecessary rebalancing.
Parity doesn't seem to have any documentation on what it's console output means. At least none that I've found which admittedly doesn't mean a whole lot. Can anyone give me a breakdown of the meaning of the following line?
2018-03-09 00:05:12 UTC Syncing #4896969 61ee…bdad 2 blk/s 508 tx/s 16 Mgas/s 645+ 1 Qed #4897616 17/25 peers 4 MiB chain 135 MiB db 42 MiB queue 5 MiB sync RPC: 0 conn, 0 req/s, 182 µs
Thanks.
Why document when you can just read code? (bleh)
2018-03-09 00:05:12 UTC(1) Syncing #4896969(2) 61ee…bdad(3) 2 blk/s(4) 508 tx/s(5) 16 Mgas/s(6) 645+(7) 1(8) Qed #4897616(9) 17/25 peers(10) 4 MiB chain(11) 135 MiB db(12) 42 MiB queue(13) 5 MiB sync(14) RPC: 0 conn(15), 0 req/s(16), 182 µs(17)
Timestamp
Best block number (latest verified block number)
Best block hash
Blocks downloaded per second
Transactions downloaded per second
Millions of gas processed per second
Unverified queue size
Verified queue size
Latest block number
Number of active peer nodes/number of total peer nodes
Blockchain header cache size
Blockchain state cache size
Queue cache size
Node sync metadata cache size
Number of open RPC sessions to your node
RPC requests per second
Approximate roundtrip ping
Now the answer to this question is also included in Parity's FAQ. It provides a comprehensive explanation of different command line output:
What does Parity's command line output mean?
Since I am still very new to Tensorflow, I don't know what caused this problem. I am currently using Tensorflow to do text classification. I needed the timing to evaluate 1 data since I need precision. When I time the evaluation time in CPU, the time required is constant while there seems to be additional time when using GPU.
Here is the statistics:
Using GPU:
1 data --> 300 ms.
10 data --> 300 ms. 30 ms each.
100 data --> 400 ms. 4 ms each.
1000 data --> 4000 ms. 4 ms each.
Using CPU :
1 data --> 10 ms.
10 data --> 100 ms. 10 ms each.
100 data --> 1000 ms. 10 ms each.
1000 data --> 10000 ms. 10 ms each.
So recently I was playing around with Django on the Jython platform and wanted to see its performance in "production". The site I tested with was just a simple return HttpResponse("Time %.2f" % time.time()) view, so no database involved.
I tried the following two combinations (measurements done with ab -c15 -n500 -k <url>, everything in Ubuntu Server 10.10 on VirtualBox):
J2EE application server (Tomcat/Glassfish), deployed WAR file
I get results like
Requests per second: 143.50 [#/sec] (mean)
[...]
Percentage of the requests served within a certain time (ms)
50% 16
66% 16
75% 16
80% 16
90% 31
95% 31
98% 641
99% 3219
100% 3219 (longest request)
Obviously, the server hangs for a few seconds once in a while, which is not acceptable. I assume it has something to do with reloading Jython because starting the jython shell takes about 3 seconds, too.
AJP serving using patched flup package (+ Apache as frontend)
Note: flup is the package used by manage.py runfcgi, I had to patch it because flup's threading/forking support doesn't seem to work on Jython (-> AJP was the only working method).
Almost the same results here, but sometimes the last 100 requests don't even get answered at all (but server process still alive).
I'm asking this on SO (instead of serverfault) because it's very Django/Jython-specific. Does anyone have experience with deploying Django sites on Jython? Is there maybe another (faster) way to serve the site? Or is it just too early to use Django on the Java platform?
So as nobody replied, I investigated a bit more and it seems like my problem might have to do with VirtualBox. Using different server OSes (Debian Squeeze, Ubuntu Server), I had similar problems. For example, with simple static file serving, I got this result from the Apache web server (on Debian):
> ab -c50 -n1000 http://ip.of.my.vm/some/static/file.css
Requests per second: 91.95 [#/sec] (mean) <--- quite impossible for static serving
[...]
Connection Times (ms)
min mean[+/-sd] median max
Connect: 0 2 22.1 0 688
Processing: 0 206 991.4 31 9188
Waiting: 0 96 401.2 16 3031
Total: 0 208 991.7 31 9203
Percentage of the requests served within a certain time (ms)
50% 31
66% 47
75% 63
80% 78
90% 156
95% 781
98% 844
99% 9141 <--- !!!!!!
100% 9203 (longest request)
This led to the conclusion that (I don't have a conclusion, but) I think the Java reloading might not be the problem here, rather the virtualization. I will try it on a real host and leave this question unanswered till then.
FOLLOWUP
Now I successfully tested a bare-bones Django site (really just the welcome page) using Jython + AJP over TCP/mod_proxy_ajp on Apache (again with patched flup package). This time on a real host (i7 920, 6 GB RAM). The result proved that my above assumption was correct and that I really should never benchmark on a virtual host again. Here's the result for the welcome page:
Document Path: /jython-test/
Document Length: 2059 bytes
Concurrency Level: 40
Time taken for tests: 24.688 seconds
Complete requests: 20000
Failed requests: 0
Write errors: 0
Keep-Alive requests: 0
Total transferred: 43640000 bytes
HTML transferred: 41180000 bytes
Requests per second: 810.11 [#/sec] (mean)
Time per request: 49.376 [ms] (mean)
Time per request: 1.234 [ms] (mean, across all concurrent requests)
Transfer rate: 1726.23 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 0 1 1.5 0 20
Processing: 2 49 16.5 44 255
Waiting: 0 48 16.5 44 255
Total: 2 49 16.5 45 256
Percentage of the requests served within a certain time (ms)
50% 45
66% 48
75% 51
80% 53
90% 69
95% 80
98% 90
99% 97
100% 256 (longest request) # <-- no multiple seconds of waiting anymore
Very promising, I would say. The only downside is that the average request time is > 40 ms whereas the development server has a mean of < 3 ms.