Microsoft MVP Logo

What in the hell has been going on with my blog recently? As I alluded on May 5th, I started having issues with my blog around April 20. For some odd reason my blog would suddenly take off and consume excessive memory. My host notified me on numerous occasions that it was taking consuming over 300MB or even 500MB. Long story short, I was shut off for the good over the other customers on the same server. I contacted others who run the same blog engine, SubText, unfortunately it seems my problem is unique. My host also moved me to different servers… but still the problem persisted.

What was so frustrating is I couldn’t duplicate the problem locally… even under stress tests that were double to my actual traffic… no matter how hard I tried, I couldn’t get the memory consumption over 90MB (still seemed a bit high). After failing numerous times to duplicate the problem locally, I’ve finally moved to a server where I have console access and can run in my own app pool in production. The good news: it seems the memory issue is still happening as I’ve got the app pool configured to recycle after a specified duration and also when the total memory reaches a certain threshold… thus my site is stay up (except for you unfortunate souls who hit me during a brief recycle). Looks like it’s happening about once an hour or two.

“That’s good news?” Actually yes, because ~finally~ I have something to work with! Now I just have to figure out how to determine exactly what is eating up all that memory. Not exactly sure how to go about this, but I’m sure I’ll have luck digging through some of Tess’ posts (who’s blog is awesome about low level debugging). Hopefully this will turn into a good case study… I’ll keep you posted.

Comments powered by Disqus