Jan 14 2007
I apologize for the site downtime today, to my huge audience of loyal readers (er… both of you). It turns out that my new ISP, TextDrive, has been having a lot of downtime on the shared hosting machine my sites are on… frequent freezes/crashes, exacerbated by long delays in noticing and recovering. Apparently it is most likely some oddball problem with some particular site belonging to one of the many customers on that machine… but unclear which one. Hopefully they will locate the trouble soon and return to the high quality of service they were previously known for.
Update: Another 1+ hour downtime on Jan. 17.
Update: Another ~1 hour downtime on Jan. 25. There were some short downtimes in the last few days also.
Update: What a surprise, another ~1 hour downtime on Jan. 27
Update: Another short downtime on Jan. 28
Update: ~2 hour downtime on Jan. 29. Seeing a pattern yet?
Update: 2.5 hour downtime on Jan. 30. This time with a kernel update.
Update: TextDrive says they have a major update coming soon, that will fix all these issues; but no specific timetable.
Update: ~1.5 hours downtime in the early overnight hours of Feb. 4. Not mentioned on the issue tracker site.
Update: ~1 hour downtime on Feb. 7, during the morning part of the workday. Maybe didn’t really want customer after all.
Update: ~1.5 hour downtime on Feb. 8, during morning part of the workday.
Update: Ah, it seems the TextDrive crew is quite busy pushing the envelope for high-end Rails hosting (via Tim Bray). It is quite impressive, but it’s also disappointing – I’d rather they had spent that brainpower making their existing hosted sites stay up. Plus… a 5-10 minutes downtime while I was trying to post this update. TextDrive considers frequent short downtimes normal, for “Apache restarts”. On several other hosts where I have had shared hosting accounts, I have not experienced this at all; so it seems to be something about how TextDrive works.
Update: After a month of stability (aside from a few short downtimes every day for “apache restarts”), ~1.5 hours downtime on March 12.
If you found this post useful, please link to it from your web site, mention it online, or mention it to a colleague.
No responses yet