Day: August 16, 2003

…although I should know better than to say that, given how things have been going around here lately.
OK, this morning’s scans completed successfully, and things appear to be functioning normally. Here’s what happened, best as I can tell:
– Last week, I had a database corruption issue, which caused me to have to drop and re-add a key table of the Ecosystem. When I did that, I forgot to re-add the indexes on the table.
– This caused the Ecosystem routines to hammer the database, and the Hosting Matters server. Apparently, HM decided to put a job in place to kill my processes. They did not, however, decide to actually notify me of this.
– I’ve now fixed the indexing problem, which hopefully will resolve the resource-usage problem for HM. I’ve emphasized to HM that I do want to be a good citizen, and am perfectly willing to work with them to not put an unreasonable load on their systems.
So was this one my fault, or Hosting Matters fault? A little of both, would be my conclusion.
Anyway, that’s the story for now; hopefully this is the end of the troubles for a while….
-NZB

…Hosting Matters. Again.
I just received confirmation that yes indeedy, Hosting Matters implemented a change a few weeks back that rendered part of my Ecosystem code inoperable.
Turns out that they have now made it illegal to execute ‘exec’ calls from php scripts, which is a crucial part of how the Ecosystem routines work. They did this change — in general terms — on their support forum, but I didn’t see it.
I’ve posted a request to the HM support forum asking for help from other HM users on how to fix the problem — as customer support has declined to offer an alternative. Suggestions from the peanut gallery here are of course most welcome as well.
I do apologize for the long delay in even attempting to fix this problem, but the demands of my non-blogging life have been high lately. It is frustrating to me as well, I assure you — I’ve put quite a bit of effort into ensuring that the Ecosystem code is reliable and keeps working as long as I don’t mess with it — beacuse I know that I don’t have time to be futzing around and fixing it every day. But it is difficult for me to react quickly to unexpected changes like this…
UPDATE: Annette from Hosting Matters has now replied to my issue, and has re-examined the problem. She’s found that ‘exec’ was not the issue, rather, the ‘jailshell’ which I believe HM defaults users to was causing the problem. It appears that I’ve now been reset to a normal, non-restricted shell.
In addition, Annette has pointed out that apparently some of my routines are not terminating properly, causing problems for HM’s server resources, and so they put a job in place to kill them for me. (I don’t blame them at all for doing so, although I’m puzzled why they didn’t contact me on the issue instead, and have indicated that in the future, I’d be happy to address this kind of problem if I’m told about it).
Anyway: I need to do some further investigation to clean up my own mess, and Annette has removed the restrictions while I do so, which seems fair. So with a little luck, I’ll have things cleaned up by the end of the weekend one way or another…
-NZB