Dietpi: META | DietPi-Benchmark optional uploads

[Fourdee]

Similar to dietpi-survey. Have an option to run:

• RootFS transfer rate
• CPU performance
• RAM performance

Then option to upload scores to http://dietpi.com/benchmark
Simple database to compare your scores against HW models etc.

@MichaIng
Thoughts?

[Fourdee]

REF: https://github.com/Fourdee/DietPi/issues/1253#issuecomment-346881878

[MichaIng]

@Fourdee
Nice idea. We can combine this with dietpi-survey, e.g. dietpi-benchmark allows user to choose upload, then creates some flag (which could be simply the data within /tmp/dietpi-benchmark/upload.txt) and starts dietpi-survey. dietpi-survey then checks for the flag and in case adds benchmark results die survey upload.

Hmm, but needs different handling, as relation between device data and results is needed 🤔.

[Fourdee]

@MichaIng

Excellent 👍

Once https://github.com/Fourdee/DietPi/issues/1768 is completed, i'll make a start on this.

[LexiconCode]

Consider implementing this into the stats for each device under the download section of dietpi.com
It's very hard to get a picture of real-world performance of a SBC before purchase not to mention accessory hardware performance in relation to a particular SBC.

[Fourdee]

@LexiconCode

Great idea, love it 👍 We'll definitely consider doing this.

My PHP experience is somewhat limited, so we may have to simply link the URL in the downloads section.

[Fourdee]

@MichaIng
Benchmark suite code done, just need to impliment how we are to upload this to dietpi.com. Use of survey and send, if the generated results file exists?

Currently, the file exported is below:

cat << _EOF_ > /var/lib/dietpi/dietpi-benchmark/upload_results
$CPUBENCH_TIME
$rootfs_write
$rootfs_read
$ram_write
$ram_read
_EOF_

root@DietPi:~# cat /var/lib/dietpi/dietpi-benchmark/upload_results
9.27
37.1
198
542
387

[LexiconCode]

Is it possible for rootfs benchmark to determine what type of hardware like MMC, hard drive, SD card?

In addition which interface it's using USB 2. 0, 3. 0, sata...

[Fourdee]

@LexiconCode

Is it possible for rootfs benchmark to determine what type of hardware like MMC, hard drive, SD card?

Unsure, however, good suggestion, i'll take a look 👍

[MichaIng]

@Fourdee
Ah sorry, was a bid busy since last weekend, missed that one.

First I agree with @LexiconCode that the values have more meaning, if connected with related hardware info.

• On SBCs CPU and RAM are fixed, thus here $G_HW_MODEL[_DESCRIPTION] would be sufficient.
• But on native PC + VM we should find out which hardware is in use, e.g. grep -m1 '^model name' <<< "$(</proc/cpuinfo)" | sed 's/model name[[:blank:]]*: //'?
• But couldn't find a way to get memory bandwidth or name info, besides using additional tools: lshw or dmidecode
• For disk speed and all machines are at least some basic drive and connection info helpful to sort/interpret the values, as mentioned. mmc can be identified via root mount /dev/* name, but no idea how to get all these infos without additional tools.
  
  
  
  • cat /sys/block/mmcblk0/device/name does not always exist
  
  
  • cat /sys/block/sda/device/model sometimes empty, also does not always exist
  
  
  • Couldn't find USB version or something so far.
  
  

About file scheme:

• If @userdeveloper98 finds the time to add it to the new survey page, theoretically it doesn't matter too much.
• But just to stay flexible it is never wrong to add it in a shell readable way again. But compared to the java database, we need to think how to present the data. E.g.:
  
  
  
  • Calculate average values for each device, also to support our performance rating. Then the values would need to be added e.g. like this:
    
    aBENCH_CPU[$G_HW_MODEL_DESCRIPTION]=$(( ${aBENCH_CPU[$G_HW_MODEL_DESCRIPTION]:=0} + $CPU_BENCH_RESULT )) && (( ${aDEVICE_COUNT[$G_HW_MODEL_DESCRIPTION]}++))
    
    After looping through all upload files:
    
    for i in ${#aDEVICE_COUNT[@]}; do ${aBENCH_CPU_AVG[$i]}=$(( ${aBENCH_CPU[$i]} / ${aDEVICE_COUNT[$i]} )); done
  
  
  • Show a result spread over all devices:
    
    ((aBENCH_CPU[$CPU_BENCH_RESULT]++))
    
    The result would need to be rounded first, grouped into reasonable score intervals, so we can present it in some sort of histogram.
  
  
• Or we do both 😄.
• But sourcing these files then does not contain any additional hardware info. For this I see no other chance then creating $UNIQUE_ID indexed arrays and adding the above mentioned infos as well to allow differentiation.

[Fourdee]

@MichaIng

We could probably do this using a 2D array?

((aBENCH_CPU_TIME[$G_HW_MODEL_DESCRIPTION,$(sed -n 1p $fp_dietpi_benchmark)]++))

That way, results are linked to the $G_HW_MODEL_DESCRIPTION

# -------------------------
# DietPi-Benchmark
# -------------------------
((aBENCH_CPU_TIME[Odroid C2 (aarch64),17.15]++))
((aBENCH_ROOTFS_WRITE[Odroid C2 (aarch64),26.6]++))
((aBENCH_ROOTFS_READ[Odroid C2 (aarch64),110]++))
((aBENCH_RAM_WRITE[Odroid C2 (aarch64),473]++))
((aBENCH_RAM_READ[Odroid C2 (aarch64),480]++))

---

I'am still unsure how you scrape the above into the survey file, but, i'd probably do it this way. Using UUID as the array index

aBENCH_DEVICE[$UUID]=$G_HW_MODEL_DESCRIPTION
aBENCH_CPU_TIME[$UUID]=$(sed -n 1p $fp_dietpi_benchmark)

---

Or even:

for (( i=0;i<total_file_count;i++))
do
. UUID $i
done

from

# -------------------------
# DietPi-Benchmark
# -------------------------
index=$1
aBENCH_DEVICE[$index]=$G_HW_MODEL_DESCRIPTION
aBENCH_CPU_TIME[$index]=$(sed -n 1p $fp_dietpi_benchmark)
aBENCH_ROOTFS_WRITE[$index]=$(sed -n 2p $fp_dietpi_benchmark)
aBENCH_ROOTFS_READ[$index]=$(sed -n 3p $fp_dietpi_benchmark)
aBENCH_RAM_WRITE[$index]=$(sed -n 4p $fp_dietpi_benchmark)
aBENCH_RAM_READ[$index]=$(sed -n 5p $fp_dietpi_benchmark)

Then we can work with the array's above.

root@DietPi:~# cat bench
# -------------------------
# DietPi-Benchmark
# -------------------------
index=$1
aBENCH_DEVICE[$index]='Odroid C2 (aarch64)'
aBENCH_CPU_TIME[$index]='17.15'
aBENCH_ROOTFS_WRITE[$index]='26.6'
aBENCH_ROOTFS_READ[$index]='110'
aBENCH_RAM_WRITE[$index]='473'
aBENCH_RAM_READ[$index]='480'
root@DietPi:~# . bench 0

[Fourdee]

Ok made a start on this code for this, basic code working well.

We end up with arrays of:

BENCH_RESULT_CPU_MIN[$HW_MODEL]
BENCH_RESULT_CPU_MAX[$HW_MODEL]
BENCH_RESULT_CPU_AVG[$HW_MODEL]

Which we can use to print some nice HTML stuff afterwards (which I have no clue where to start lol)

root@DietPi:/var/www/downloads/testing# ./dietpi-survey_report
#system 1
17.09
72.1
110
468
489

#system 2 (copy of system 1 upload)
17.09
72.1
110
468
489

BENCH_RESULT_CPU_AVG for HW_MODEL 12 = 17.0

Temps not printing once added to array, but the original var is?

[Fourdee]

@MichaIng

Whoop whoop 🥂

[MichaIng]

@Fourdee
Niiiice.

Maybe we can merge the results as additional columns into the "Devices" install count table, what you think?

[Fourdee]

@MichaIng

Maybe we can merge the results as additional columns into the "Devices" install count table, what you think?

Good idea 👍

Only issue is, i'll need to create about 4/5 of these tables, as, adding all the info on 1 table makes it too squashed.

I'll have a play bit later.

---

Also really want it this way, but can't figure it out yet:

[Fourdee]

Some issues with server, needs a reboot:

[10535890.095824] TCP: request_sock_TCP: Possible SYN flooding on port 443. Sending cookies.  Check SNMP counters.
[10538980.248728] TCP: request_sock_TCP: Possible SYN flooding on port 80. Sending cookies.  Check SNMP counters.
[12131781.922070] INFO: task jbd2/sda1-8:410 blocked for more than 120 seconds.
[12131781.922074]       Not tainted 4.9.0-6-amd64 #1 Debian 4.9.82-1+deb9u3
[12131781.922074] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[12131781.922076] jbd2/sda1-8     D    0   410      2 0x00000000
[12131781.922098]  ffff8c6cbacac000 0000000000000000 ffff8c6c769030c0 ffff8c6cbfc18940
[12131781.922124]  ffffffff9da11500 ffffabcbc04b7ca0 ffffffff9d40c649 ffff8c6c76e7e888
[12131781.922126]  0000000000000246 ffff8c6cbfc18940 ffffabcbc04b7d80 ffff8c6c769030c0
[12131781.922129] Call Trace:
[12131781.922252]  [<ffffffff9d40c649>] ? __schedule+0x239/0x6f0
[12131781.922269]  [<ffffffff9ceb9c50>] ? prepare_to_wait_event+0xf0/0xf0
[12131781.922271]  [<ffffffff9d40cb32>] ? schedule+0x32/0x80
[12131781.922326]  [<ffffffffc02f428f>] ? jbd2_journal_commit_transaction+0x25f/0x17b0 [jbd2]
[12131781.922329]  [<ffffffff9ceb9c50>] ? prepare_to_wait_event+0xf0/0xf0
[12131781.922333]  [<ffffffffc02f9c02>] ? kjournald2+0xc2/0x260 [jbd2]
[12131781.922335]  [<ffffffff9ceb9c50>] ? prepare_to_wait_event+0xf0/0xf0
[12131781.922338]  [<ffffffffc02f9b40>] ? commit_timeout+0x10/0x10 [jbd2]
[12131781.922346]  [<ffffffff9ce970c9>] ? kthread+0xd9/0xf0
[12131781.922349]  [<ffffffff9ce96ff0>] ? kthread_park+0x60/0x60
[12131781.922355]  [<ffffffff9ce7c3d0>] ? SyS_exit_group+0x10/0x10
[12131781.922357]  [<ffffffff9d411537>] ? ret_from_fork+0x57/0x70
[12131781.922370] INFO: task mysqld:24660 blocked for more than 120 seconds.
[12131781.922371]       Not tainted 4.9.0-6-amd64 #1 Debian 4.9.82-1+deb9u3
[12131781.922372] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.

root@DietPi:/var/www/downloads/testing# netstat -s
Ip:
    Forwarding: 2
    2092417752 total packets received
    14006058 with invalid addresses
    0 forwarded
    26 with unknown protocol
    0 incoming packets discarded
    2059706761 incoming packets delivered
    1870713670 requests sent out
    40 outgoing packets dropped
    673460 dropped because of missing route
    182 reassemblies required
    91 packets reassembled ok
    26 fragments failed
Icmp:
    66821 ICMP messages received
    2453 input ICMP message failed
    InCsumErrors: 39
    ICMP input histogram:
        destination unreachable: 41795
        timeout in transit: 1538
        echo requests: 23377
        echo replies: 72
    99260 ICMP messages sent
    0 ICMP messages failed
    ICMP output histogram:
        destination unreachable: 75884
        echo replies: 23376
IcmpMsg:
        InType0: 72
        InType3: 41795
        InType8: 23377
        InType11: 1538
        OutType0: 23376
        OutType3: 75884
Tcp:
    5589296 active connection openings
    37983973 passive connection openings
    145804 failed connection attempts
    5531654 connection resets received
    7 connections established
    2081679398 segments received
    5236087781 segments sent out
    52396075 segments retransmitted
    33289 bad segments received
    7526591 resets sent
    InCsumErrors: 32707
Udp:
    400670 packets received
    77623 packets to unknown port received
    113 packet receive errors
    391943 packets sent
    0 receive buffer errors
    0 send buffer errors
    InCsumErrors: 113
    IgnoredMulti: 20273762
UdpLite:
TcpExt:
    1074141 SYN cookies sent
    9125102 SYN cookies received
    723124 invalid SYN cookies received
    145004 resets received for embryonic SYN_RECV sockets
    246 ICMP packets dropped because they were out-of-window
    8656124 TCP sockets finished time wait in fast timer
    1215248 packetes rejected in established connections because of timestamp
    3210833 delayed acks sent
    19825 delayed acks further delayed because of locked socket
    Quick ack mode was activated 468480 times
    14330436 times the listen queue of a socket overflowed
    14940709 SYNs to LISTEN sockets dropped
    44418 packets directly queued to recvmsg prequeue
    276408 bytes directly in process context from backlog
    TCPDirectCopyFromPrequeue: 13783
    52279651 packet headers predicted
    155 packet headers predicted and directly queued to user
    926133001 acknowledgments not containing data payload received
    787114555 predicted acknowledgments
    4366 times recovered from packet loss due to fast retransmit
    TCPSackRecovery: 5262837
    TCPSACKReneging: 2904
    Detected reordering 44248 times using FACK
    Detected reordering 1845103 times using SACK
    Detected reordering 1592 times using reno fast retransmit
    Detected reordering 152248 times using time stamp
    34303 congestion windows fully recovered without slow start
    124914 congestion windows partially recovered using Hoe heuristic
    TCPDSACKUndo: 117757
    722787 congestion windows recovered without slow start after partial ack
    TCPLostRetransmit: 980038
    710 timeouts after reno fast retransmit
    TCPSackFailures: 917323
    239687 timeouts in loss state
    39121399 fast retransmits
    3053065 forward retransmits
    2621607 retransmits in slow start
    TCPTimeouts: 5765772
    TCPLossProbes: 1769261
    TCPLossProbeRecovery: 145940
    TCPRenoRecoveryFail: 740
    TCPSackRecoveryFail: 347299
    TCPDSACKOldSent: 457720
    TCPDSACKOfoSent: 4058
    TCPDSACKRecv: 13748494
    TCPDSACKOfoRecv: 258584
    5301177 connections reset due to unexpected data
    72478 connections reset due to early user close
    188155 connections aborted due to timeout
    3 times unable to send RST due to no memory
    TCPSACKDiscard: 17370
    TCPDSACKIgnoredOld: 507873
    TCPDSACKIgnoredNoUndo: 1323291
    TCPSpuriousRTOs: 680006
    TCPSackShifted: 144424895
    TCPSackMerged: 87101372
    TCPSackShiftFallback: 64126953
    TCPDeferAcceptDrop: 37059644
    TCPReqQFullDoCookies: 1958200
    TCPRetransFail: 3783
    TCPRcvCoalesce: 10550775
    TCPOFOQueue: 1136028
    TCPOFOMerge: 2984
    TCPChallengeACK: 17615
    TCPSYNChallenge: 1582
    TCPFastOpenCookieReqd: 3503
    TCPAutoCorking: 2010
    TCPFromZeroWindowAdv: 2
    TCPToZeroWindowAdv: 2
    TCPWantZeroWindowAdv: 8
    TCPSynRetrans: 3062891
    TCPOrigDataSent: 5098952431
    TCPHystartTrainDetect: 27286
    TCPHystartTrainCwnd: 5349343
    TCPHystartDelayDetect: 394714
    TCPHystartDelayCwnd: 33381101
    TCPACKSkippedSynRecv: 4515
    TCPACKSkippedPAWS: 1073376
    TCPACKSkippedSeq: 59730
    TCPACKSkippedFinWait2: 229
    TCPACKSkippedTimeWait: 2729
    TCPACKSkippedChallenge: 19627
    TCPWinProbe: 697466
    TCPKeepAlive: 108
IpExt:
    InMcastPkts: 5
    InBcastPkts: 20283769
    InOctets: 165045175955
    OutOctets: 7222217202331
    InMcastOctets: 160
    InBcastOctets: 1172424085
    InNoECTPkts: 2090911592
    InECT1Pkts: 64073
    InECT0Pkts: 1536471
    InCEPkts: 1353760

[Fourdee]

@MichaIng

Server offline for backup.

Might also upgrade it to 4cores 4GB 😉 | wont let me upgrade, ticket sent. We can pay the difference (£5 + month)

[MichaIng]

@Fourdee

Also really want it this way, but can't figure it out yet:

Use colspan to spread a cell over multiple columns. rowspan could be used to spread the first "Device" cell over the first two lines then:

<table>
  <tr>
    <th rowspan="2">Device:</th>
    <th colspan="3">CPU time</th>
  </tr>
  <tr>
    <th>avg</td>
    <th>min</td>
    <th>max</td>
  </tr>
  <tr>
    <td>RPi</td>
    <td>123</td>
    <td>56</td>
    <td>203</td>
  </tr>
</table>

---

Server offline for backup.

Hmm Cloudflare's Always Online does not work here 🤔.

[Fourdee]

Boom boom:

[MichaIng]

@Fourdee
Looks nice.

However this could be presented much nicer on the new java survey page with nice charts.
Maybe @userdeveloper98 finds some time to implement this by times? 😜

[Fourdee]

Color me up sally:

[Fourdee]

@MichaIng

However this could be presented much nicer on the new java survey page with nice charts.
Maybe @userdeveloper98 finds some time to implement this by times? 😜

Yep agree, we'll need to wait and see if @userdeveloper98 is up for it :)

[Fourdee]

Getting alot of this during update:
Was occurring before any of my changes.

TOSTART_OPTION[]: bad array subscript
/tmp/dietpi-survey_report/11f218f5-2ba7-4835-af54-b3e1da5f2e91.txt: line 17: aSOFTWARE[${aSOFTWARE_NAME6_10[]}]++: bad substitution

---

line 19: aSOFTWARE[]: bad array subscript

#!/bin/bash
# -------------------------
((aSURVEY_VERSION[6]++))
# -------------------------
((aSURVEY_SENTCOUNT[5]++))
((aDIETPI_VERSION[6.9]++))
((aDEVICE_NAME[RPi 3 Model B (armv7l)]++))
((aCPU_ARCH[armv7l]++))
((aCPU_COUNT[4]++))
((aDISTRO_VERSION[stretch]++))
((aAUTOSTART_OPTION[${aAUTOSTART_NAME[5]}]++))
((aAUTO_SETUP_AUTOMATED[0]++))
((aNETWORK_INTERFACE[eth0]++))
# -------------------------
# DietPi-Software Installed
# -------------------------
((aSOFTWARE[${aSOFTWARE_NAME6_9[5]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[6]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[31]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[44]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[96]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[103]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[104]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_9[152]}]++))

---

line 17: aSOFTWARE[${aSOFTWARE_NAME6_10[]}]++: bad substitution

#!/bin/bash
# -------------------------
((aSURVEY_VERSION[6]++))
# -------------------------
((aSURVEY_SENTCOUNT[1]++))
((aDIETPI_VERSION[6.10]++))
((aDEVICE_NAME[NanoPi NEO 2 (aarch64)]++))
((aCPU_ARCH[aarch64]++))
((aCPU_COUNT[4]++))
((aDISTRO_VERSION[stretch]++))
((aAUTOSTART_OPTION[${aAUTOSTART_NAME[0]}]++))
((aAUTO_SETUP_AUTOMATED[0]++))
((aNETWORK_INTERFACE[eth0]++))
# -------------------------
# DietPi-Software Installed
# -------------------------
((aSOFTWARE[${aSOFTWARE_NAME6_10[]}]++))

lol?

Ok we need to check min/max.

((aAUTOSTART_OPTION[${aAUTOSTART_NAME[28]}]++))

[MichaIng]

@Fourdee
Jep known issues.

On earlier versions for it was possible that survey added an empty string as installed software ID. I guess it was due to first run survey, if not .installed file was created yet, or at least nothing yet marked as installed.

The other issue is due to DietPi subversion integer being wrong sometimes in updater. So DietPi is actually v6.10+, thus survey uploads new survey file scheme, but version string added is wrong, so no special aSOFTWARE_NAME array exist for this.

I anyway though that we could clean the survey files a bid, removing those that were not changed for time X. But not urgent, since the errors above do not break anything, just a few software installs are dropped.

[Fourdee]

6103e95d-3a09-4729-b7dd-6bd9f3cb2e7c.txt: line 17: aSOFTWARE[]: bad array subscript
((aSOFTWARE[${aSOFTWARE_NAME_12[5]}]++))

[Fourdee]

@MichaIng

since the errors above do not break anything

Adding checks for bad files and skipping them 👍

root@DietPi:/var/www/downloads/testing# /etc/cron.minutely/dietpi-survey_report
Bad file contains array with no index ([]), skipping: /tmp/dietpi-survey_report/00fa6266-cc32-410c-9017-df2ce5003b85.txt

since the errors above do not break anything

~Might possibly be the cause~ why only 1 benchmark result is showing, i've done 4, unsure yet:
https://dietpi.com/survey/#benchmark

Appears, the upload completes, but file does not exist on server.

[  OK  ] DietPi-Survey | Connection test: ssh.dietpi.com
#!/bin/bash
# -------------------------
((aSURVEY_VERSION[6]++))
# -------------------------
((aSURVEY_SENTCOUNT[23]++))
((aDIETPI_VERSION[6.17]++))
((aDEVICE_NAME[RPi 3 Model B+ (armv7l)]++))
((aCPU_ARCH[armv7l]++))
((aCPU_COUNT[4]++))
((aDISTRO_VERSION[stretch]++))
((aAUTOSTART_OPTION[${aAUTOSTART_NAME[0]}]++))
((aAUTO_SETUP_AUTOMATED[0]++))
((aNETWORK_INTERFACE[eth0]++))
# -------------------------
# DietPi-Software Installed
# -------------------------
((aSOFTWARE[${aSOFTWARE_NAME6_17[0]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_17[103]}]++))
((aSOFTWARE[${aSOFTWARE_NAME6_17[105]}]++))
# -------------------------
# DietPi-Benchmark
# -------------------------
BENCH_HW_MODEL=3
BENCH_CPU=21.30
BENCH_ROOTFS_WRITE=7.2
BENCH_ROOTFS_READ=20.0
BENCH_RAM_WRITE=506
BENCH_RAM_READ=159
BENCH_CPU_TEMP_START=44
BENCH_CPU_TEMP_END=59

[ .... ] DietPi-Survey | (1/5) Testing connection to ssh.dietpi.com, please wait[  OK  ] DietPi-Survey | Connection test: ssh.dietpi.com
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   830    0     0  100   830      0    812  0:00:01  0:00:01 --:--:--   812
100   830    0     0  100   830      0    812  0:00:01  0:00:01 --:--:--   812
[  OK  ] DietPi-Survey | Successfully sent survey data

root@DietPi:/var/www/downloads/testing# sync
root@DietPi:/var/www/downloads/testing# nano /home/dietpi-survey/survey/dce3cbc9-d3c5-4506-a654-c475184786bd.txt
#no file

Aha
http://www.manpagez.com/man/5/sshd_config/
MaxSessions

Idiot moment, resolved below 🈯️

[Fourdee]

Completed, all issues resolved 🈯️

You can run a quick benchmark and upload with:

/DietPi/dietpi/func/dietpi-benchmark 2

View results:
https://dietpi.com/survey#benchmark

[MichaIng]

@Fourdee
What do you think about creating results file in a shell source-able way, so it can be directly cat into the dietpi-survey upload file and sourced by other scripts as well. Generally I like to do this for most of our data/settings/results files, as sourcing should always be much faster then using sed or grep to catch content. Also the same values are then forced to have the same variable names across scripts, to have increased consistency.
https://github.com/Fourdee/DietPi/blob/dev/dietpi/func/dietpi-benchmark#L253-L261

[Fourdee]

@MichaIng

Yep, makes sense, i'll make the change.

[Fourdee]

@MichaIng

Done, cleaned up and now sourced vars.

Tested fine.

[MichaIng]

@Fourdee
Currently upload is done automatically, so no "optional uploads". Can we have a switch to have results saved locally or upload them? Or two separate selections for bench only and upload only?

I am thinking as well, if it is okay to bench single devices as well, e.g. CPU only, but then we need to handle empty entries in report.

And currently on VM, 1° CPU temperature is shown. Would be good to not only show N/A on report page, but as well in dietpi-benchmark menu header/results window. At best the temp test is simple skipped then in case of empty values, show N/A.

[Fourdee]

@MichaIng

I am thinking as well, if it is okay to bench single devices as well, e.g. CPU only, but then we need to handle empty entries in report.

The whole purpose of this was to reduce available options, and, support benchmark uploads via 1 menu item.

I removed the option to manually run CPU/rootfs/RAM as its contained within the suite option, which, we want users to run as share/view their results.

Currently upload is done automatically, so no "optional uploads"

Yep, intended, its part of the survey. However, we should change the "Ok" to run bench, have it a "Yes/No":

And currently on VM, 1° CPU temperature is shown.

Yep, i'll see if I can clean it up.

[Fourdee]

🈯️

[Fourdee]

Completed.

[MichaIng]

@Fourdee

The whole purpose of this was to reduce available options, and, support benchmark uploads via 1 menu item.

Hmm, I still think we should allow to run benchmarks without upload. Even that our survey is anonym and safe, I don't like it from end user perspective to be forced to transfer any data. Nothing against reducing options for simplicity, but there should be always a choice, if it's about sharing your data.
E.g. I purged free version of 3dmark instantly from my Windows, after recognizing that bench results are automatically uploaded 😉.

[Fourdee]

@MichaIng

Thought the same eariler.

~2 tests:~

Ok so:

• Opted in = automatic upload of scores
• Opted out = Prompt to opt in, if user still opted out, no upload, however, scores are now shown after test.

I'll make the change.

[Fourdee]

Completed.