HFM.NET - Monitoring Application for Folding@Home v7

[bruce]

If you configure FAH so that the data files are in dropbox each version of the files will be backed up every time they're updated -- but it will probably work.

A VPN such as Hamchi would be another way of doing it.

A third method would be to open a connection through your firewall and you share FAH's data files. You can then mount them on the other computer using normal network functions.

Since the files are updated frequently, any of these methods will add a continuous low-level bit of network traffic. Moreover you'd be creating new security risks. (You may be able to manage those security risks -- I don't know enough about your knowledge of internet security to judge.) Both Hamchi and Dropbox have one level of built-in security which would not be present with a network share method until you added it.

I've used Hamchi.

[DrBB1]

Thanks for the suggestion, compdewd. I have done this but will need to wait a couple of days until my next next WU can start. Will report back to share findings.

I followed compdewd's suggestion, but no luck. I'll wait for the next revision and see if that takes care of the issue.

[PantherX]

Does anyone know or have suggestions on how to merge two WuHistory.db3 files into one so that HFM.NET can still read it?

[compdewd]

Does anyone know or have suggestions on how to merge two WuHistory.db3 files into one so that HFM.NET can still read it?

I imagine you'd have to use an application that can create a new database file from scratch and add the two WuHistory.db3 files by importation. AFAIK the file just has to be in SQLite database format for HFM.NET.

[GreyWhiskers]

Does anyone know or have suggestions on how to merge two WuHistory.db3 files into one so that HFM.NET can still read it?

Excellent question. I have 4 computers running F@H, all of them also running HFM. I got back a few weeks ago from a 6-week long road trip. I had my laptop with me, and folded whenever we were stable in a hotel room. Also left the other computers back at home folding. Of course, my travel laptop had the history of only itself while on the road, and my home computers had only them.

I'd really like to combine the WuHistory files also so I can get a more comprehensive view now that I'm back at home.

[k1wi]

So long as you save backup copies of them I think you're best off taking compdewd's advice. You might have to investigate the database to ensure there aren't any conflicts when merging (for example if there is a unique ID for each entry).

[PantherX]

I have two instances of HFM.NET. One on my laptop only and the other for all the desktops. To edit the WuHistory.db3 files, I use the following:

SQLite Administrator -> I found the GUI not that intutive when compared to some others. Moreover, I couldn't figure out how to select multiple rows of data via the GUI (Shift key didn't work).

SQLite Database Browser -> Initially, it kept crashing when I tried to navigate my download folder. I got around that problem by copying my WuHistory.db3 file to a different location. You can't select multiple records (Shift key didn't work). Moreover, it did crash but fortunately, it didn't destroy the database.

SQLite Manager -> Firefox Add-on that has a nice GUI, can select multiple records via the GUI (Shift key works). This is what I would recommend to use if any modifications to WuHistory.db3 needs to be carried out.

I still can't figure out how to merge the two WuHistory.db3 files together. Moreover, there is a problem with the fact that the sequence numbers are repeated across the two database and they are the primary key which means that duplicates aren't allowed. Will see what I can do (or others suggest) and will post the (hopefully positive instead of negative) results here.

[k1wi]

Possibly the easiest way to get around the primary key issue would be to take the maximum sequence number of one db and then add that value to the other db in a new field (i.e. so row '1' gains a new field with the value 1+db_n), then set it as the primary key and drop the existing primary key field... Probably a good idea to only do that on back-up db files though...That's how I'd go about such a method, but I haven't used sqlite.

Stackoverflow might give you some sqlite specific advice.

[PantherX]

Thanks for the tip K1wi. Will work on it and hopefully, will have something to share.

[jimerickson]

just a note to let you know i got HFM running and producing a web site on the Raspberry Pi. had to use Mono 2.11.4 hard float experimental version to get it to run. it is available here https://www.dropbox.com/s/sask17flot3zq ... binary.tgz for anyone interested. do note that you must remove any other version of mono for this to work. this is a cheap remote monitoring solution for those of us running multiple machines.

[GreyWhiskers]

Observation: Duplicate HFM WU History Entries

I've seen a spate lately of duplicate WU history entries - roughly correlated to a spate of upload/download issues that may be my network carrier (Comcast), or maybe due to the Stanford servers, or may be due to some of the characteristics of the current (7.3.6) FAH software.

I'm using HFM software 0.9.1, the latest available.

I detail here one simple instance- only two entries for the completed work unit. I have other examples where there were many times repeated unsuccessful upload attempts. One has 33 HFM WU History entries over the 36 minutes of repeated upload attempts.

I still use the HFM history data via the SQL database for various EXCEL analysis, and thankfully it is easy for Excel to be programmed to eliminate duplicates. But, HFM doesn't seem to be triggering on the "Upload Complete" "final Credit Estimate" lines. This may be due to the fact that these events are entered in the FAH log but are not passed across the Client Socket Interface and official APIs.

When I have upload issues and the upload has to start multiple times, I am getting multiple line items in the HFM log. I can't completely correlate them - but let me give an example from both the HFM logs and the FAH logs. I've shortened what I post here to what I think is the most relevant, but obviously more is available.

This instance is for completed Work Unit project:8090 run:628 clone:0 gen:8 core running in Folding Slot 1 (SMP5) on my computer named DigiStorm. Slots 0 and 2 are two GPUs also folding on this machine.

The FAH log is in GMT; the HFM log seems to be in PDT local time.

FAH shows the following Upload attempts (highly edited log transcript - full transcript below):

- 16:15:18:WU03:FS01:Connecting to 171.67.108.60:8080
- 16:15:50:WU03:FS01:Upload 66.20%
- 16:15:50:WARNING:WU03:FS01:Exception: Failed to send results to work server: Transfer failed

- 16:15:50:WU03:FS01:Trying to send results to collection server
- 16:15:50:WU03:FS01:Uploading 4.15MiB to 171.65.103.160
- 16:15:50:WU03:FS01:Connecting to 171.65.103.160:8080
- 16:16:21:ERROR:WU03:FS01:Exception: Transfer failed

- 16:16:22:WU03:FS01:Sending unit results: id:03 state:SEND error:NO_ERROR project:8090 run:628 clone:0 gen:8 core:0xa4 unit:0x0000000b6652edcc51674aa8dab63eb9
- 16:16:22:WU03:FS01:Uploading 4.15MiB to 171.67.108.60
- 16:16:22:WU03:FS01:Connecting to 171.67.108.60:8080
- 16:17:02:WARNING:WU03:FS01:Exception: Failed to send results to work server: Transfer failed

- 16:17:02:WU03:FS01:Trying to send results to collection server
- 16:17:02:WU03:FS01:Uploading 4.15MiB to 171.65.103.160
- 16:17:02:WU03:FS01:Connecting to 171.65.103.160:8080
- 16:17:52:ERROR:WU03:FS01:Exception: Transfer failed

- 16:17:52:WU03:FS01:Sending unit results: id:03 state:SEND error:NO_ERROR project:8090 run:628 clone:0 gen:8 core:0xa4 unit:0x0000000b6652edcc51674aa8dab63eb9
- 16:17:52:WU03:FS01:Uploading 4.15MiB to 171.67.108.60
- 16:17:52:WU03:FS01:Connecting to 171.67.108.60:8080
- 16:18:32:WU03:FS01:Upload complete
- 16:18:32:WU03:FS01:Server responded WORK_ACK (400)
- 16:18:32:WU03:FS01:Final credit estimate, 7093.00 points

Now, HFM ended up with two WU history entries:

[5/14/2013-09:15:19 AM] - (DigiStorm Slot 01) Writing unit to database: P8090 (R628, C0, G8)
[5/14/2013-09:15:50 AM] - (DigiStorm Slot 01) Writing unit to database: P8090 (R628, C0, G8)

These are entries I copied from the HFM Work Unit History display. Both entries are at 9:15 AM - this log display doesn't show seconds.
8090 p8090 DigiStorm Slot 01 Unknown GRO_A4 2.27 00:06:48 0.75 7,107.8 15,051.9 5/13/2013 09:40 PM 5/14/2013 09:15 AM 100 FINISHED_UNIT 39000 628 0 8

8090 p8090 DigiStorm Slot 01 Unknown GRO_A4 2.27 00:06:48 0.75 7,110.4 15,057.3 5/13/2013 09:40 PM 5/14/2013 09:15 AM 100 FINISHED_UNIT 39000 628 0 8

FAH V7.3.6 log of repeated upload failures until one finally got through. Filtered for time, and to show only Folding Slot 01 - the SMP slot.

Code: Select all

16:15:18:WU03:FS01:0xa4:Folding@home Core Shutdown: FINISHED_UNIT
16:15:18:WU03:FS01:FahCore returned: FINISHED_UNIT (100 = 0x64)
16:15:18:WU03:FS01:Sending unit results: id:03 state:SEND error:NO_ERROR project:8090 run:628 clone:0 gen:8 core:0xa4 unit:0x0000000b6652edcc51674aa8dab63eb9
16:15:18:WU03:FS01:Uploading 4.15MiB to 171.67.108.60
16:15:18:WU03:FS01:Connecting to 171.67.108.60:8080
16:15:18:WU02:FS01:Starting
16:15:18:WU02:FS01:Running FahCore: "C:\Program Files (x86)\FAHClient/FAHCoreWrapper.exe" C:/Users/Al/AppData/Roaming/FAHClient/cores/www.stanford.edu/~pande/Win32/AMD64/beta/Core_a3.fah/FahCore_a3.exe -dir 02 -suffix 01 -version 703 -lifeline 5052 -checkpoint 15 -np 5
16:15:18:WU02:FS01:Started FahCore on PID 4524
16:15:18:WU02:FS01:Core PID:6284
16:15:18:WU02:FS01:FahCore 0xa3 started
16:15:19:WU02:FS01:0xa3:
16:15:19:WU02:FS01:0xa3:*------------------------------*
16:15:19:WU02:FS01:0xa3:Folding@Home Gromacs SMP Core
16:15:19:WU02:FS01:0xa3:Version 2.27 (Dec. 15, 2010)
16:15:19:WU02:FS01:0xa3:
16:15:19:WU02:FS01:0xa3:Preparing to commence simulation
16:15:19:WU02:FS01:0xa3:- Looking at optimizations...
16:15:19:WU02:FS01:0xa3:- Created dyn
16:15:19:WU02:FS01:0xa3:- Files status OK
16:15:19:WU02:FS01:0xa3:- Expanded 1948785 -> 2929924 (decompressed 150.3 percent)
16:15:19:WU02:FS01:0xa3:Called DecompressByteArray: compressed_data_size=1948785 data_size=2929924, decompressed_data_size=2929924 diff=0
16:15:19:WU02:FS01:0xa3:- Digital signature verified
16:15:19:WU02:FS01:0xa3:
16:15:19:WU02:FS01:0xa3:Project: 7513 (Run 0, Clone 23, Gen 85)
16:15:19:WU02:FS01:0xa3:
16:15:19:WU02:FS01:0xa3:Assembly optimizations on if available.
16:15:19:WU02:FS01:0xa3:Entering M.D.
16:15:24:WU03:FS01:Upload 45.13%
16:15:25:WU02:FS01:0xa3:Mapping NT from 5 to 5 
16:15:25:WU02:FS01:0xa3:Completed 0 out of 500000 steps  (0%)
16:15:30:WU03:FS01:Upload 51.15%
16:15:37:WU00:FS00:0x17:Completed 375000 out of 2500000 steps (15%)
16:15:50:WU03:FS01:Upload 66.20%
16:15:50:WARNING:WU03:FS01:Exception: Failed to send results to work server: Transfer failed
16:15:50:WU03:FS01:Trying to send results to collection server
16:15:50:WU03:FS01:Uploading 4.15MiB to 171.65.103.160
16:15:50:WU03:FS01:Connecting to 171.65.103.160:8080
16:15:58:WU03:FS01:Upload 6.02%
16:16:21:WU03:FS01:Upload 42.12%
16:16:21:ERROR:WU03:FS01:Exception: Transfer failed
16:16:22:WU03:FS01:Sending unit results: id:03 state:SEND error:NO_ERROR project:8090 run:628 clone:0 gen:8 core:0xa4 unit:0x0000000b6652edcc51674aa8dab63eb9
16:16:22:WU03:FS01:Uploading 4.15MiB to 171.67.108.60
16:16:22:WU03:FS01:Connecting to 171.67.108.60:8080
16:16:28:WU03:FS01:Upload 4.51%
16:16:41:WU03:FS01:Upload 9.03%
16:17:02:WU03:FS01:Upload 24.07%
16:17:02:WARNING:WU03:FS01:Exception: Failed to send results to work server: Transfer failed
16:17:02:WU03:FS01:Trying to send results to collection server
16:17:02:WU03:FS01:Uploading 4.15MiB to 171.65.103.160
16:17:02:WU03:FS01:Connecting to 171.65.103.160:8080
16:17:10:WU03:FS01:Upload 1.50%
16:17:21:WU03:FS01:Upload 4.51%
16:17:33:WU03:FS01:Upload 37.61%
16:17:52:WU03:FS01:Upload 39.12%
16:17:52:ERROR:WU03:FS01:Exception: Transfer failed
16:17:52:WU03:FS01:Sending unit results: id:03 state:SEND error:NO_ERROR project:8090 run:628 clone:0 gen:8 core:0xa4 unit:0x0000000b6652edcc51674aa8dab63eb9
16:17:52:WU03:FS01:Uploading 4.15MiB to 171.67.108.60
16:17:52:WU03:FS01:Connecting to 171.67.108.60:8080
16:18:02:WU03:FS01:Upload 1.50%
16:18:22:WU03:FS01:Upload 13.54%
16:18:28:WU03:FS01:Upload 70.71%
16:18:32:WU03:FS01:Upload complete
16:18:32:WU03:FS01:Server responded WORK_ACK (400)
16:18:32:WU03:FS01:Final credit estimate, 7093.00 points
16:18:32:WU03:FS01:Cleaning up

HFM Log - edited to the same time frame as the FAH log, and filtered for only the entries from my computer named DigiStorm (this HFM instance was monitoring four separate local computers running FAH V7)

Code: Select all

[5/14/2013-01:31:15 AM] - 
[5/14/2013-01:31:15 AM] - Starting - HFM.NET v0.9.1.595
[5/14/2013-01:31:15 AM] - 
[5/14/2013-01:31:28 AM] - SQLite DLL Path: C:\Program Files (x86)\HFM.NET\SQLite\AMD64\System.Data.SQLite.dll

[5/14/2013-09:15:05 AM] - (DigiStorm Slot 01) Slot Status: Running
[5/14/2013-09:15:19 AM] ! (DigiStorm Slot 01) Could not find log section for slot 1. Cannot update frame data for this slot.
[5/14/2013-09:15:19 AM] - (DigiStorm Slot 01) Writing unit to database: P8090 (R628, C0, G8)
[5/14/2013-09:15:20 AM] - (DigiStorm Slot 01) Slot Status: Running
[5/14/2013-09:15:50 AM] - (DigiStorm Slot 00) Slot Status: Running
[5/14/2013-09:15:50 AM] - (DigiStorm Slot 02) Slot Status: Running
[5/14/2013-09:15:50 AM] - (DigiStorm Slot 01) Writing unit to database: P8090 (R628, C0, G8)
[5/14/2013-09:15:50 AM] - (DigiStorm Slot 01) Slot Status: RunningNoFrameTimes
[5/14/2013-09:15:50 AM] - (DigiStorm) Retrieval finished in 114 ms
[5/14/2013-09:16:20 AM] - (DigiStorm Slot 00) Slot Status: Running
[5/14/2013-09:16:20 AM] - (DigiStorm Slot 02) Slot Status: Running
[5/14/2013-09:16:20 AM] - (DigiStorm Slot 01) Slot Status: RunningNoFrameTimes
[5/14/2013-09:16:20 AM] - (DigiStorm) Retrieval finished in 61 ms
[5/14/2013-09:16:50 AM] - (DigiStorm Slot 00) Slot Status: Running
[5/14/2013-09:16:50 AM] - (DigiStorm Slot 02) Slot Status: Running
[5/14/2013-09:16:50 AM] - (DigiStorm Slot 01) Slot Status: RunningNoFrameTimes
[5/14/2013-09:16:50 AM] - (DigiStorm) Retrieval finished in 59 ms
[5/14/2013-09:17:20 AM] - (DigiStorm Slot 00) Slot Status: Running
[5/14/2013-09:17:20 AM] - (DigiStorm Slot 02) Slot Status: Running
[5/14/2013-09:17:20 AM] - (DigiStorm Slot 01) Slot Status: RunningNoFrameTimes
[5/14/2013-09:17:20 AM] - (DigiStorm) Retrieval finished in 85 ms
[5/14/2013-09:17:50 AM] - (DigiStorm Slot 00) Slot Status: Running
[5/14/2013-09:17:50 AM] - (DigiStorm Slot 02) Slot Status: Running
[5/14/2013-09:17:50 AM] - (DigiStorm Slot 01) Slot Status: RunningNoFrameTimes
[5/14/2013-09:17:50 AM] - (DigiStorm) Retrieval finished in 80 ms
[5/14/2013-09:18:20 AM] - (DigiStorm Slot 00) Slot Status: Running
[5/14/2013-09:18:20 AM] - (DigiStorm Slot 02) Slot Status: Running
[5/14/2013-09:18:20 AM] - (DigiStorm Slot 01) Slot Status: RunningNoFrameTimes
[5/14/2013-09:18:20 AM] - (DigiStorm) Retrieval finished in 61 ms

[HaloJones]

Nothing clever but I've been playing with the webpage output templates to produce this:

http://www.miketimbers.webspace.virginm ... mmary.html

Anyone else care to share theirs?

[PantherX]

I notice that for me, I get multiple entries if it takes longer to upload the WU (in addition to the failed attempts). I had instances of 92 duplicates that I manually removed. I do hope that in the next version, this can be resolved.

[bruce]

As I'm sure you know, we can't really provide any support for HFM. Has anybody contacted Harlam directly about this problem? He doesn't visit this forum as often as he once did.

[GreyWhiskers]

I just sent Harlam a note -- let's see what comes up.

Thanks for the reminder, Bruce.