A side note:

In order to speed up returning results, I use 0.0 and 0.05 for the values of getting work and 'additional' work -
except when i set my device profile for hst and for for 0.8 days of additional work.
I reset the work times and device profile for open Zika (and/or other projects) in the morning after the HST fishing times..

So far, I haven't found a statistic in WCG for WU turnaround time (including idle time that the WU waits to be processed in my PC.)

Does anyone know If this exists somewhere?

Thanks,
Jay

PS
for example, Einstein has this:
Average turnaround time: 0.07 days

Sandvika, as far as I know - to BOINC, WCG is just 1 project. The WCG projects (Zika, HST, FAH2, MCM etc.), are all "sub projects" within the 1 WCG project - thus, I'll be surprised to find out that it's possible to only request an update for HST WU's.

Obviously, I could well be proved wrong - so please don't take my word as "gospel".

My procedure is similar to jay_Orlando's except that I change the profile on the server, not the local profile. That is easier if you have several computers at different locations. I am also using BoincTasks to trigger a refresh of all. The strange thing is that only one of 5 computers has gotten a decent supply of HSTB, enough that it has been crunching only HSTB for 2 weeks. I still have a 2-day supply of HSTB, which might be enough to get me to gold.

Cheers

Having recently reached sapphire on HSTB, I can say it's a real effort to crunch this project. It's admittedly anecdotal but from what I have seen, it would appear that I rarely get HSTB WUs on the weekends. I do not see the 03/53 pattern I've read about. My suspicion is that projects normally have an automated task that loads new work into the servers and that that task runs on some sort of automated schedule but for HSTB, that automated task is only running on weekdays when it can be easily monitored by the admins. Whatever the case, I do not recall seeing more than a dozen HSTB WUs on any single machine of mine and that has been extremely rare. Usually, no more that 4-6 is the best I see. Trying to make good progress on this project with just a few machines is very challenging. I have 25 machine running a total of 46 threads. 20 machines are Ubuntu Linux with Xeon cpus. The other five machines are Wintels - 3 WIN10, 1 Vista and 1 XP. My approach has been one that probably is too much on micro-managing for most but has worked for me as I am normally on my machines each evening anyway. I run all of them with one single profile normally set with HSTB only selected. I have set my cache to 4 days. That is normally much more than necessary for other projects. WCG limits machines to 35 WU's per core/thread so a 4 day cache usually means that almost all of my machine hit the 35 per thread limit before reaching the cache limit. On the Linux machines, 35 WUs is typically 1.5 - 2 days of work. Two of the WIN10 machines will reach the 4 day cache limit as they are 8 threads each. The other machines fall in between. When I am on my machines in the evening, I modify the server profile to add ZIKA along with HSTB. That allows my machines to "fill up" on work ensuring that they don't run dry. After 5-10 minutes, I'll change the server profile back to HSTB only. This approach keeps my machine crunching but normally they are almost always asking the WCG servers for HSTB work. What you get for the projects you select always depends on what is available at the time your request hits the WCG servers so my approach does tend to "nag" the servers for work. The comparatively short run times for ZIKA wu's means more frequent requests for new work and seems to improve the odds. I usually average having one HSTB wu per thread across all my machines. I do not like having to "nag" the servers for work as that is fruitless more often than not but it seems necessary for this project - at least until the supply improves by multiple orders of magnitude. As I am nearing my goal for ZIKA, I will soon be switching from using ZIKA to back fill to FAAH and/or FAH2. Since the run time for those wu's is significantly greater, I expect that I will have to make major changes to my approach to crunching HSTB. This approach has allowed me to pass sapphire in the 9-10 months this project has been out but that has also been with 46 threads running 24x7 the whole time.

Yep. With the Zika currently loaded on my machines, I'll go a couple of months past ten years. I have some FAAH loaded and they are ranging from 11 minutes on my Xeons to 45 minutes on my slowest machine. That means most of my machines will only have a few HOURS of work cached before they hit the 35 per thread/core limit. As I recall, FAH2 WUs were quite long so that will change the picture yet again once we start getting FAH2 again. In the meantime, I finish up my remaining ZIKA over the next couple of days or so. It's going to be interesting, but probably sad, to see what HSTB I get while I'm running FAAH since I won't be able to ask for just HSTB.