MySQL Database & Cron Job Failures

If you run tt-rss on an officially unsupported platform (shared hosting, Windows, etc.) post here
SOV
Bear Rating Trainee
Bear Rating Trainee
Posts: 15
Joined: 13 Jun 2014, 02:41

MySQL Database & Cron Job Failures

Postby SOV » 15 Jun 2014, 06:28

No issues currently, just wanting to know how a few things work.

Setup:
1.) Shared Hosting: DreamHost
2.) MySQL Database
3.) 85 Feeds
4.) I use the "Reeder" app for viewing articles, the app caches all information I need. I hardly use the TRSS web portal from my website.


Questions:
1.) From what I've read the database size doesn't really matter, its the number of queries so, is stability based off of the number of times I refresh the feeds or the amount of articles per feed?

2.) Is there any risk in losing articles if the database size gets too large or if a Cron Job fails? I have a Cron Job set up every 15 minutes and I usually get a temporary failure twice a day, doesn't seem to be a big issue.

3.) I have read wiki articles here and some forum posts but I couldn't quite gather exactly how purging works. I know it is necessary to prevent duplicate articles but is there a more in depth explanation anywhere? I'm just curious.

4.) Considering I can't install PostgreSQL, is there anything I can do to help stabilize TRSS for the future, considering I have a lot of feeds and articles coming in daily? GZIP is already enabled by default on DreamHost's servers.

JustAMacUser
Bear Rating Overlord
Bear Rating Overlord
Posts: 373
Joined: 20 Aug 2013, 23:13

Re: MySQL Database & Cron Job Failures

Postby JustAMacUser » 16 Jun 2014, 01:39


SOV
Bear Rating Trainee
Bear Rating Trainee
Posts: 15
Joined: 13 Jun 2014, 02:41

Re: MySQL Database & Cron Job Failures

Postby SOV » 16 Jun 2014, 03:10

Thank you so much for this reply. I definitely appreciate your time in typing that response. A lot of information there for me.

1.) I am using the Fever plugin by the way to access TT-RSS through Reeder. Seems like I’ve stayed on top of my RSS Feeds since installation and I’ve not noticed the 50 article limit per request. The iOS Reeder App enables local background refresh on my phone so it randomly powers up to refresh when my phone is plugged into a power source.

I double checked my database and all the tables are set to the “InnoDB” type. I read one person on the forum a couple of his tables were “MyISAM” so I’m glad that issue didn’t happen to me.

2.) I looked at the error.log and the only errors I seen is a client [IP ADDRESS] looking for two files, one saying robots.txt and missing.html at the root directory. I have no idea why that’s there, but other than that no issues. The one Cron Job error I was getting usually twice a day was “setlock: fatal: unable to lock /tmp/cronlock.3786865475.165183: temporary failure”. I found an article on DreamHost’s website and someone recommended unchecking the “Use Locking” option in the Cron Job tab of DreamHost’s website. I just unchecked that option and I’ll see if any other error pursues. The descriptor states “Use locking: Prevents this job from running more than once at a time. This is generally a good thing!”. So if anything it looks like if there’s a temporary failure this selection will let it run again, instead of waiting another 15 minutes.

3.) What would you consider a huge database size?

4.) I guess I’m using the wrong term “to stabilize”. I guess I really mean optimize. Since TT-RSS isn’t supported on shared hosting I guess I had this feeling of wanting to double check settings and just make sure there wasn’t anything else I could do to make it run smoothly or without issue. I didn’t want to take for granted that everything is running smooth currently thinking I missed something.

Again, thanks for your reply.

JustAMacUser
Bear Rating Overlord
Bear Rating Overlord
Posts: 373
Joined: 20 Aug 2013, 23:13

Re: MySQL Database & Cron Job Failures

Postby JustAMacUser » 16 Jun 2014, 05:10


Bob
Bear Rating Trainee
Bear Rating Trainee
Posts: 7
Joined: 22 Jun 2013, 03:08

Re: MySQL Database & Cron Job Failures

Postby Bob » 17 Jun 2014, 00:17

TT RSS runs fine on Dreamhost, currently I have about 20000 (i know, it was 70000 - but i deleted a few feeds) items. Updates set to run every 5 minutes (cron) and feeds between 20 minutes and 1 hour (so the cron isn't always making tt-rss do a full update)

The cron error is normal, you checked the only allow on instance in the cron manager on the DH web panel. and only appears because the previous cron job is still running.

SOV
Bear Rating Trainee
Bear Rating Trainee
Posts: 15
Joined: 13 Jun 2014, 02:41

Re: MySQL Database & Cron Job Failures

Postby SOV » 17 Jun 2014, 05:56


JustAMacUser
Bear Rating Overlord
Bear Rating Overlord
Posts: 373
Joined: 20 Aug 2013, 23:13

Re: MySQL Database & Cron Job Failures

Postby JustAMacUser » 17 Jun 2014, 06:52

It think he's saying that in the DreamHost control panel he sets the cron job to run in 5-minute intervals, but in TT-RSS the feeds refresh at most every 20-minutes. This means there would be fewer updates each time the cron job is run.

SOV
Bear Rating Trainee
Bear Rating Trainee
Posts: 15
Joined: 13 Jun 2014, 02:41

Re: MySQL Database & Cron Job Failures

Postby SOV » 17 Jun 2014, 07:07

Ohh I see. Part of me being confused was because I assumed the Feed refreshing time in the TTRSS panel was redundant if you had to use a Cron Job to refresh your feeds.

I thought a 10 minute Cron Job 'does' refresh your feeds.

I thought the refresh in the TTRSS panel was your settings if you had background processes enabled on a dedicated server.

I thought once you set a Cron Job it overrides the panel refresh options.


Sent from my iPhone using Tapatalk

JustAMacUser
Bear Rating Overlord
Bear Rating Overlord
Posts: 373
Joined: 20 Aug 2013, 23:13

Re: MySQL Database & Cron Job Failures

Postby JustAMacUser » 18 Jun 2014, 03:59

You set how often you want each feed to be updated in TT-RSS. When the cron or daemon run, they'll check which feeds have not been updated within the set interval and then update those. So the two work hand in hand. Running the daemon keeps a process open so it "spawns" a child process to check about every two minutes. With cron jobs, the cron schedule is what spawns the child process (so to speak), so every 5-30 minutes is good.


Return to “Unsupported platforms”

Who is online

Users browsing this forum: No registered users and 1 guest