User Discussions > Welcome & Notices

TM data, logs, stats, etc. archives

(1/1)

Jim:
For about the last year, the site update process has archived the previous state of TM as part of each update.  I had pretty much forgotten it was keeping things for so long.  It takes a lot of storage (about 357 GB at the moment) but the server has no shortage of disk space at this point, so I plan to keep saving things.  So if you ever want to have access to an old log file, old stats about the contents of a highway system, etc., you can send along a request by email to me or in a PM here on the forum.   Saved for each site update are the .sql file used to populate the DB, and the contents of the graphdata, lots, nmp_merged, and stats directories.

Also from these archives, tools like "grep" can easily generate fun output like one attached which shows my overall stats in active+preview systems over the last year.  Nice to see I've had a productive year, adding 3000 miles between new travels and previous travels on systems that have reached preview or active status in TM.

yakra:
I remember digging into these files a bit when troubleshooting... something, during C++ development.

Out of curiosity, what was the grep command you used? I see just a partial line, of the relevant stats sans the "Overall in active+preview systems:" bit.

Jim:
I cheated.  I used


--- Code: ---grep "Overall in active+preview" */logs/users/terescoj.log
--- End code ---

but then edited out the extra redundant bits in the file.

Edit: this one would give the output I posted:


--- Code: ---grep "Overall in active+preview" */logs/users/terescoj.log | sed "s=/logs/users/terescoj.log:Overall in active+preview systems==g"

--- End code ---

michih:

--- Quote from: Jim on May 27, 2019, 03:43:54 pm ---It takes a lot of storage (about 357 GB at the moment) but the server has no shortage of disk space at this point, so I plan to keep saving things.
--- End quote ---

I'd love seeing historical mileage progress on the site one day :)
I think that it would be enough to store data of the first* update each year or each month (or week?) if there would be any shortage one day...

*Which would represent the last update of the previous year/month/week.

yakra:

--- Quote from: Jim on May 27, 2019, 03:43:54 pm ---Saved for each site update are the .sql file used to populate the DB, and the contents of the graphdata, lots, nmp_merged, and stats directories.

--- End quote ---
This just came in handy, as it turns out. I just grabbed one of the .sql files from when this issue was live, and installed it on lab2. It seems MySQL 5.7 & MySQL 8.0 differ in some pretty significant ways on how they handle multi-byte characters and over-length fields.

Navigation

[0] Message Index